Archive for December, 2010


Crustless Cherry Berry Tarts With Olive Oil Frangipane (Baked)

[ Equipment: oven or slow cooker with temperature control. For more information about the terminology in this recipe, see Low Temperature Baking: A Journey of 3 Paths ].

The term crustless refers to the lack of a pastry crust. These tarts are all “filling”, the fruits embedded in a leavened pastry cream batter. The pictures above show 2 versions of the crustless tart. The top one has the lighter batter, cherries and delicate top crust. The bottom tart is denser, firmer and packed with blueberries. They are variations on Rachel Allen’s No-Pastry Pear and Almond Tart, which she demonstrated in an episode of her TV show (see a picture of Allen’s tart here).

Clafoutis may be the most famous no-pastry tart, but her tart was not a clafoutis, which has a wetter batter at nearly 4 times the level of hydration (compared with Julia Child’s clafoutis recipe). Rather, the batter’s ingredient roster suggested a frangipane, an almond pastry cream, but formulated with a higher ratio of flour to almond meal and leavened with aerated egg whites to create a texture and body closer to that of a cake.

In my version of this recipe, I kept the Allen’s flour-to-almond-meal ratio (about 1:1 by volume), but changed out the butter for a combination of olive oil and milk. My experience has been that butter tends to dry out LTB cakes. As to the choice of oil, I thought the fruitiness of an extra virgin olive oil made it a fine substitute for butter in this recipe, adding both fat and flavor to the batter and blending nicely with the cherries and blueberries. Unlike conventional baking, LTB will preserve the taste of the olive oil. The oil’s only a partial substitution for the full amount of butter to keep the calorie count down, although at 410 calories per tart, it’s not diet food. Milk covers the remainder for hydration.

I also experimented with the amount and preparation of the egg whites and the addition of egg yolks. In Allen’s recipe, the egg whites were beaten for 30 seconds until frothy and then mixed into the frangipane batter. The air in the whites, invigorated by a 400°F/200°C oven, blew up the tart. However, an LTB oven at 250°F/121°C doesn’t develop that kind of puffing heat, so the volume of the raw whites pretty much sets the height. With that limitation in mind, I tried whites beaten to the foam stage (stabilized with lemon juice) and beaten to the stiff peak stage. With the stiff-peak whites, the tart baked up like a cottony chiffon cake, but tasted and felt like a light almond cream as it dissolved in the mouth. The tart with the foamed whites came out moist and dense, like a traditional frangipane.

Both types are described in this recipe. Type 1 has the stiff-peak egg white and cherry fruit. Type 2 incorporates the foamed egg whites and blueberries. Although the overall hydration is about the same in both recipes, the type 2 has less milk (and twice as much egg white). I recommend a superfine or baker’s sugar for the type 2, because it will dissolve faster or more thoroughly in the reduced hydration of batter prior to folding in the egg whites.

What did I do with the leftover egg yolks? Well, I saved them for other things like custard-base ice milks and enriched scrambled eggs. I did try a type 1 tart with a yolk (shown above on the left). It tasted too eggy, too much like a cake, really overpowering the fruit. On the other hand, the yolk tart held its shape better, no cracks in the top crust, and with a denser texture, like a type 2 tart. In the type 1 recipe, I list an optional amount of ground flax seed to get some of the cohesion of an egg yolk without its standout flavor.

Makes 2 mini tarts
– 410 calories per tart
– Oven Temperature: 250°F/121°C

Type 1 (stiff-peak egg white batter):

  • 1/4 cup almond meal
  • 1/4 cup all-purpose flour
  • 1/4 cup sugar
  • 1/8 cup extra virgin olive oil
  • 1/4 cup milk
  • 1/4 teaspoon almond extract
  • 1/8 teaspoon ground flax seed (optional, see text)
  • 1 egg white
  • 1 tablespoon sugar
  • pinch of salt
  • 1/8 teaspoon lemon or lime juice
  • 8 cherries, pitted and sliced lengthwise in half

Type 2 (foamed egg white batter):

  • 1/4 cup almond meal
  • 1/4 cup all-purpose flour
  • 1/4 cup sugar
  • 1/8 cup extra virgin olive oil
  • 1/8 cup milk
  • 1/4 teaspoon almond extract
  • 2 egg whites
  • 1/8 teaspoon lemon or lime juice
  • pinch of salt
  • 1/4 cup fresh blueberries

1. Grease 2 4-inch mini tart pans and set aside. The non-stick pans above were greased and floured, but did not release much better than pans that were greased only.

2. In a medium bowl, sift flour and almond meal. Mix in the sugar and combine thoroughly. Instead of sifting, I will put the flour, almond meal and sugar in the bowl and rub it between my hands to combine them and eliminate any lumps.

3. Type 1 tart: in another bowl, whip the egg white with the tablespoon of sugar, lemon juice and salt until it reaches the stiff peak stage.

Type 2 tart: in another bowl, whip the egg whites with the lemon juice and salt until it becomes a light foam. A fork is perfectly adequate for this job.

4. Add milk and olive oil to the flour mixture and whisk until well combined. Note: the type 2 batter will be much thicker in this step because it has less milk.

5. Type 1 tart: Fold the stiffened egg white into the batter.

Type 2 tart: Gently mix the foamed egg whites into the batter.

6. Fill each tart pan about 3/4 full of batter.

7. Place cherries cut-side up in batter or sprinkle blueberries into batter. The cherry tart on the left is a type 1. The one with blueberries on the right is a type 2.

8. Bake for 65 to 75 minutes, until the edges are golden brown. Remove to a rack and cool. Note that type 1 tarts will sink and crack as they cool.

9. Unmold and serve. As seen in the first picture above, the top crust on the type 1 (cherry) tart breaks apart easily. I saved the pieces and reassembled them on the tart once it was unmolded. I garnished the cherry tart with chopped walnuts.


Shim Mod For Black&Decker CBM210/220 Burr-Mill Grinder

This article documents how I modified a Black & Decker CBM210 burr-mill grinder to be an acceptable flour mill. When I began baking gluten-free and low-carb recipes, I knew I would have to grind some of my own flours. The only grinding apparatus I owned were an immersion blender and a rotary blade coffee grinder. Neither could produce fine quality flour in large quantities. I wasn’t prepared to invest in a flour mill, but I had heard that less-expensive burr-mill coffee grinders could produce flour fine enough for baking. Locally, the only affordable burr-mill grinder was the Black & Decker CBM210 (the CBM220 – identical in appearance – adds automatic features). It retailed for about $30 and had been on sale for as low as $20 (cheaper than the smaller CBM205).

I purchased a 210 without reading any reviews online and later found out that this model is notorious for inconsistently grinding coffee beans (coarse grounds mixed with fine powder on all settings). My 210 performed no better, the gritty flour submitted to the grinder multiple times before I could consider it marginally acceptable. For me, there was no alternative. It was grind flour with the 210 or limit myself to commercial flours sold in my local markets.

On the product page at, several reviews describe how to shim the 210 to push the grinding wheels closer together for a finer, more consistent grind. One section of the grinding assembly (that center knob inside the hopper) can be removed and taken part to separate the grinding wheel. Shims cut from card stock (such as a business card) or thin plastic (such as a yogurt cup lid) go under the grinding wheel before reassembling. The mod transforms the 210 from a throw-away to a functional burr-mill grinder, they said.

Because I wash the grinding assembly after making flour, I didn’t want a shim cut out from card stock. Although I had several types of flat sheet plastic from which I could cut a shim (blister packs, old library cards, document protectors), most of them were too thick. I chose the lid from a storage container for my silicone cupcake molds. That lid was molded from a very thin, flexible, clear plastic. I don’t know the exact thickness, but I believe that it was about the same or actually a little thinner than the average paper-based business card. It’s better to err on the side of caution in the selection of the shim material. If the plastic is too thick, the shim could press the wheels hard against each other, completely stripping the burrs and ruining them.

The picture above shows the grinding assembly before and after modification. Before the shim mod, the wheel on the left is slightly worn along the top edge from making a batch or two of flour. The wear indicates that the wheel isn’t true or flat. On close examination of the grinding assembly, the wheel can be seen to sit off-kilter in the assembly due to poor manufacturing tolerances. On the right, with the shim in place and after grinding a batch of flour, the wear pattern has grown in a dramatic arc, streaking along half the circumference of the wheel.

Despite the increased wear, the mod does work. My grinder now produces a finer, more consistently textured flour. The flour is still grittier than a commercial flour, but can be remedied with a soaked-grain technique. The picture above shows a chocolate snack cake made with an oat-rice flour ground in the modified CBM210 mill. I soaked the flour for 1 hour in an acidulated liquid before adding the remaining ingredients. The soaking softened the grit for a finer cake than even one made with a store-bought Bob’s Red Mill Gluten-Free Flour.

Stacking a second shim in the mill might produce a finer flour but risks ruining the grinding wheel altogether.

Of course, this mod voids the warranty. I recommend it only if the grinder would otherwise be tossed in the garbage. The increased wear on the wheel means that it will have to be replaced sooner. The mod applies to the CBM210, but I have read that the CBM220 is identical except for the inclusion of automatic features that won’t be affected by the mod.

The 210 and 220 have been available since 2008 at least, and could be discontinued at any time. Thus, an article about the shim mod might, in the long term, have only limited value, and I thought about whether to write it at all. I wrote it first for the owners of these machines and second in case the performance of other similar grinders might benefit from this kind of shim mod.

The Modification

1. Remove the grinding assembly from the hopper by twisting the knob to the right until it releases. See the CBM210/220 manual for more instructions about this procedure.

2. Remove the two screws underneath the grinding assembly. Mark the screws and the orientation of the parts so that they can be reassembled in the original order. I put a strip of masking tape on the parts to mark the orientation. The inset shows the inside of the retainer that holds the wheel.

3. With a marker that writes on plastic, trace the outside and inside circumference of the wheel into the plastic. Also mark the positions of the screw holes.

4. With a craft knife, cut out the center hole.

5. With a craft drill installed with a 1/8-inch bit, drill out the screw holes. Only 2 screws attach the wheel to the grinding assembly. I drilled the extra holes for ventilation and draining when the assembly is rinsed under the faucet. Without those extra holes, water could build up behind the wheel.

6. With scissors, finish cutting out the shim along the outside circumference. Test fit the shim in the retainer and trim off any plastic as necessary. If either screw hole isn’t perfectly centered, GENTLY reform it by using the drill as a reamer.

7. Reassemble the grinding assembly with the new shim under the wheel. Do not tighten the screws excessively. The mod is now complete. Test grind a small batch of flour (or espresso if the machine is a dedicated coffee grinder) on the finest setting and check the wheel for any excessive wear. Discard the test batch.


Papaya-Chocolate-Filled Krispy Bonbon Cookies (Dehydrated)

[ Equipment: Food dehydrator or dehydration oven. For more information about the terminology in this recipe, see Low Temperature Baking: A Journey of 3 Paths ].

The inspiration for these cookies was the classic Rice Krispies Treats. They’re molded and baked in a dehydration oven for a dry, crisp texture, very different from the sticky, gooey, greasy feel of old-time, pan-cut Treats squares. Ground almonds give body to the porous rice cereal cookie dough. The center filling is a fused sandwich of papaya and chocolate, a less common yet very delicious mix of flavors. Chocolate piping dresses the cookies up for company.

To make the filling, thin strips of dried papaya are wrapped around a chocolate chip, a variation of the technique for making the date-wrapped chips in my Black Sugar Trifecta cookies. As the cookies bake, the heat melts the chocolate into a layer between the papaya. Fruit leather strips would be an excellent substitute for the papaya. In fact, any number of solid fillings would be excellent in these cookies. A moister filling, such as a mix of peanut butter and chips or peanut butter and nuts, could work. Fillings that are too moist might impede the drying process. Of the fillings I’ve tried, I made one batch filled with a 1/4 teaspoon of mini chocolate chips, and another batch filled with date-wrapped chips. Candied nuts are up next on my checklist.

The standard substitute for marshmallow creme is marshmallows melted with a small amount of corn syrup, but there are recipes that allow for a direct substitution of marshmallows for marshmallow creme. Since marshmallows are solid at room temperature, a marshmallow binder might set too quickly. As an alternative to corn syrup, a pat of butter melted with the marshmallows might help lengthen the set time. I don’t give exact amounts here because I haven’t made this substitution yet.

Makes 8 cookies
– 70 calories per cookie
– Oven Temperature: 120-130°F/49-54°C in a dehydrator


  • 1 cup puffed rice cereal
  • 4 tablespoons marshmallow creme
  • 1/2 tablespoon light agave syrup
  • 1/8 teaspoon almond extract
  • 1/8 teaspoon vanilla extract
  • 1/4 cup finely ground almonds or almond meal

Papaya-Chocolate Filling

  • 2 dried papaya spears
  • 8 regular size chocolate chips

Chocolate Glaze

  • 1 oz. Hershey’s Special Dark chocolate bar

Chocolate-Papaya Filling Method:

1. Slice papaya spears into 8 strips about 3/4 x 1-1/2 x 1/16 (inch) strips.

2. Press the pointy tip of a chocolate chip into one end of a strip.

3. Fold the other end of the papaya strip over the chip to form a pocket and press down on the edges to seal the pocket.

4. Repeat steps 2 and 3 for the remaining chips.

Chocolate Glaze Method:

Note: Prepare the melted chocolate after the cookies have finished baking and cooled.

1. Chill chocolate bar until hard and break or slice off 1 oz. of chocolate. The picture above shows 1 oz. of pieces from a Hershey’s Special Dark Chocolate bar.

2. Slice or chop chocolate into slivers and place in heatproof dish.

3. In a microwave, heat chocolate on HIGH power for 20 to 30 seconds and stir. Repeat heat-stir until chocolate is melted and smooth (about 2 minutes). Alternatively, melt the chopped chocolate in a double boiler.

4. Fold a square of wax paper (about 6 x 6 inches) into a small piping bag and fill with melted chocolate. The bag should have a tip hole that extrudes a 1/8-inch line of chocolate. If necessary, enlarge the hole by snipping the tip with scissors.

Cookies Method:

Note: The cookies shown below were baked in my homemade Presto Chango dehydration oven. The instructions may need to be modified to work with commercial food dehydrators.

1. Put ground almonds, marshmallow creme, vanilla and almond extracts, and the agave syrup in a heatproof bowl.

2. Microwave the mixture on HIGH for about 30 seconds until the marshmallow creme has just liquified. Stir to combine ingredients.

3. Add puffed rice cereal and gently stir until evenly coated with the marshmallow creme.

4. With a greased spoon, fill half of a greased 1/8-cup-capacity ice cream scoop with the puffed rice, pressing a slight indentation in the center.

5. Place a  papaya-wrapped chip in the center.

6. Fill the scoop to level with more puffed rice mix. With the back of the greased spoon, gently press down on the puffed rice to compact it (without crushing the rice).

7. Release the cookie from the scoop onto a greased dehydrator liner. The one in the picture was cut from wax paper, 8 inches in diameter with a 1/2-inch center hole to facilitate heat flow.

8. Repeat steps 4 to 7 to make a total of 8 cookies.

9. Transfer the cookies (with liner) to the dehydrator. Dehydrate at 120°F/49°C for 6 to 8 hours or until the cookies feel dry to the touch and hold their shape when lifted up. A higher dehydration temperature will reduce drying time, but risk damaging the enzymes in the raw ingredients.

10. Let cool thoroughly. If desired, pipe or spread melted chocolate over cookies.


Harvest-Spiced Oat-Rice Soda Bread (Gluten-Free, Baked)

[ Equipment: oven or slow cooker with temperature control, a 5-3/4 x 3 x 2 (inch) loaf pan. For more information about the terminology in this recipe, see Low Temperature Baking: A Journey of 3 Paths ].

“Harvest spice” is my name for regular pumpkin pie spice. It can enhance a range of baked goods from muffins to breads, many without a hint of pumpkin in them, so I call it a harvest spice in reference to the fall season when pumpkins are brought to market. In this gluten-free soda bread with milk-plumped raisins, it adds a note of autumn familiarity. The bread itself is a variation of a basic gluten-free soda bread recipe I found on The wet batter and low temperature baking created something more kin to a dessert bread. It lacks a true crust. At one time, I considered dabbing an alkaline solution on top to see if I could get it to brown (the same browning agent applied on my wet dough focaccia) but postponed that experiment for another day. Still, it’s a very tasty dessert bread.

This bread was my first foray into slow baking with a flour other than wheat. The original recipe was made with rice flour and tapioca starch. However, I wanted to try grinding my own flour, and homemade rice flour famously turns out baked goods with a gritty texture if it’s the main ingredient in the flour blend. Tapioca starch is still a specialty item and not sold in markets near me.

Early on, I tested soda breads with a commercial gluten-free flour: Bob’s Red Mill Gluten-Free All Purpose Flour, but the taste of the flour was too distinctive, possibly because of the large proportion of beans (garbanzo and fava). From the reviews on, preferences for this product are highly individual. For my own flour, I went with a classic blend: oats and white rice – with a 2-to-1 emphasis on the oat flour. (For those who don’t want to grind their own flour, many supermarkets sell oat flour and rice flour. Just blend them in a 2 (oats) to 1 (rice) ratio for this recipe.)

Any coffee grinder can convert the soft oats into a fine flour, but not so with raw rice. A regular blade coffee grinder smashes raw rice and other hard grains and legumes into a pebbly powder, too coarse for baking. Further, a blade grinder is inefficient as a flour mill, when large quantities of flour are required. On the other hand, burr-mill coffee grinders can work as flour mills at a fraction of the cost of a flour mill.

I purchased a Black & Decker burr-mill grinder (model CBM210) that costs a few dollars more than a regular blade grinder. The reviews on warned that this inexpensive machine would not produce an espresso grind (ultra-fine) consistently, and they were correct. My solution was to pass the flour through the mill several times, each pass outputting a finer, less gritty powder. Even then, the flour was never as fine as the Bob’s Red Mill flour. (To improve the performance of this machine, see my Shim Mod For Black&Decker CBM210/220 Burr-Mill Grinder.)

I chose an oats-rice ratio of 2 to 1 to create a light flour. It weighs about 10% less than the same volume of the Bob’s Red Mill flour. My rationale was that lighter flours rise faster and with less leavening, two important considerations in low temperature baking. I did try to grind a blend containing beans, but the flour performed inconsistently, possibly because the bean particles were too large and weighed down the batter. Perhaps I could have cycled the flour through the grinder more times to pulverize it more thoroughly. Another option was to sift out the larger flour particles with a fine sieve for reprocessing.

Initially, I made the bread without the ground flax seed. The slices were so delicate that they would crumble resting in my hand. More egg might have firmed up the bread at the risk of turning it into a cake. An alternative (and vegan) binder, flax seed has become a common ingredient in gluten free baking. My first batters with ground flax seed contained 1/4 to 1/2 teaspoon of seed – not an unusual amount of binder for the quantity of flour – and refused to rise. It seems slow ovens don’t generate adequate lift from chemical leaveners to counter the gelatinizing effect of the flax.

A little ground flax seed goes a long way. I found that 1/8 teaspoon of ground flax seed balanced a firm texture with a small sacrifice in volume. In the recipe, I recommend as little as 1/16 teaspoon of flax for a fluffier loaf that still holds itself together. Ground flax seed can spoil rapidly, so I make a new batch every few months (I grind the seeds in my blade coffee grinder) and store it in a jar until I’m actually making the recipe (that is, I don’t make an oat-rice-flax flour). If flax seed isn’t available, try substituting with ground chia seeds or a bit of xantham gum.

Flours that aren’t uniformly fine ground will produce baked goods with a gritty texture, but even commercial gluten-free flours may have a subtle grittiness inherent in the ingredients. I could, for example, detect the hint of bumpiness in breads made with the Bob’s Red Mill Gluten-Free General Purpose Flour, and that product is probably as fine ground as any flour in the market. I tried soaking flours for several hours in buttermilk to soften the texture (the soaking technique allegedly helps release important nutrients trapped in foods as well), and discovered that soaking impedes the rise, because the individual grains absorb water and weigh more and because some grains turn gummy. Now, if I bake with a gritty flour, I wrap the cooled loaf in plastic and let it sit for several hours before serving. The grains will absorb moisture and soften. The rest period doesn’t eliminate gritty particles, just softens them so they’re less objectionable.

As with quick breads in general, work quickly and don’t over-mix. There’s no gluten to toughen the bread from over-mixing, but overworked batter might turn gummy and fail to rise. I mixed up some loaves with the standard puree attachment on my immersion blender in an attempt to break up any large flour particles, and the batters didn’t perform well at all (little rise or no rise). Also, the baking soda begins bubbling as soon as it contacts the buttermilk. If the batter isn’t baked immediately, that soda lift could be lost.

On the topic of baking soda, the standard pairing seems to be 1/2 teaspoon baking soda per cup of buttermilk. My recipe specifies twice that ratio. The amount of soda must balance the acid in the buttermilk, which may vary from brand to brand. If the bread tastes a little bitter and has a dryish, coarse texture, try reducing the soda to 1/8 teaspoon.

Makes 6 servings
– 80 calories per serving
– Oven Temperature: 250°F/121°C

Soda Bread

  • 5/8 cup 2:1 oat-rice flour, commercial or homemade (2.8 oz – see below, see text)
  • 1/4 cup cultured buttermilk
  • 1 tablespoon beaten egg
  • 2 tablespoons milk
  • 1 tablespoon sugar
  • 1/8 teaspoon salt
  • 1/4 teaspoon harvest spice blend (see below and see text)
  • 1/16 to 1/8 teaspoon ground flax seed (see text)
  • 1/4 teaspoon baking soda (see text)
  • 1-1/4 teaspoon baking powder
  • 1 tablespoon soaked, chopped raisins (see below)

Oat-Rice Flour (makes 2-1/4 cups flour)

  • 1 cup long grain white rice
  • 2 cups rolled oats

Milk Infused Raisins (makes about 1 tablespoon)

  • 1 tablespoon raisins
  • 1/4 cup hot milk

Harvest Spice Blend (makes about 1 teaspoon)

  • 1/2 teaspoon powdered cinnamon
  • 1/4 teaspoon powdered ginger
  • 1/4 teaspoon nutmeg
  • 1/8 teaspoon cloves

Oat-Rice Flour Method

Note: The oat-rice flour for this recipe consists of 2 parts oat flour to 1 part rice flour. These grinding instructions apply specifically to the Black and Decker CBM210 Burr-Mill Grinder. They may need to be modified to work with other model grinders.

1. Cover rice with water and soak for 30 minutes.

2. Drain rice, pat dry with paper towels. Spread rice on a large baking sheet over more paper towels and allow to air dry completely or place rice in a dehydration oven set to 120°F/49°C and dry for several hours (stir the rice occasionally to help release the vapor).

3. Grind rice one or more times until fine as flour. In my burr-mill machine, I grind the rice 1 time on the coarse setting and then 3 times on the finest setting.

3. Measure out 3/4 cup of rice flour and place in a large bowl. Repeat step 2 for the oats. Measure out 1-1/2 cups of oat flour and mix thoroughly with rice flour.

5. Grind oat-rice flour blend 2 to 3 times on finest setting. If a flour sieve is available, sift the flour to filter out the larger particles and regrind them.

4. Store oat-rice flour in an airtight container.

Milk-Infused Raisins Method:

1. In a small bowl, soak raisins in hot milk for about 30 minutes.

2. Drain raisins and chop.

Soda Bread Method:

1. In a small cup, mix the sugar, salt, harvest spice blend, ground flax seed. Set aside.

2. Cut a strip of aluminum foil 6 inches wide and at least 11 inches long. Press the foil into a small loaf pan (5-3/4 x 3 x 2 inches) to cover the bottom and the long sides. Grease the foil and exposed sides of pan. Set aside.

3. In a medium bowl, mix the oat-rice flour, buttermilk, milk and beaten egg.

4. Pour in the sugar mixture and quickly whisk until combined.

5. Add baking soda and baking powder and quickly whisk until combined.

4. Stir in chopped raisins.

6. Pour batter into the loaf pan.

7. Grease the shiny side of a piece of aluminum foil about 9 x 6 inches. Cover pan with the foil (shiny side down) and crimp it securely down on and under the rim. With a 1/8-inch thick skewer, punch 5 ventilation holes along the center line, spaced about 1 inch apart.

8. Bake for 35 to 40 minutes. One way to test for readiness is to stick a toothpick into the loaf through the center ventilation hole. If it comes out clean, the bread is ready.

9. Cool for 10 minutes in the pan. Then loosen the loaf by sliding a knife along the two ends of the pan. Grasp the flaps of the foil liner and slowly pull up to lift the bread out. Remove the foil liner and continue cooling on a rack.

10. Slice and serve. Store the bread wrapped in the refrigerator (the milk-infused raisins will spoil at room temperature). If the bread has a gritty texture, allow the grains to soften by refrigerating the loaf for several hours or overnight. Reheat before serving.