soil test getting the samples right

I decided (finally) to do some soil tests with logan labs. Now my soil is build up over the years, when I moved in there was a thin thin layer of poor soil over fill. Now my soil is considerably better but I still can't grow certain things well like beetroot or brassicas. The problem is that I add to every bed compost or manure etc as I go. So most likely all beds are different. I decided that it is maybe a good idea to take one test in a bed which did not grow something and another bed which does better, maybe two or three. What do you think?

hau Angelika,  your idea is great, I've listed some of my methodology below for you should you want to use it.

The best method would be to sample each bed but that is going to get expensive if you have a lot of beds.

What I would suggest (what I do myself) is first labeling beds that are fairly similar in contents (I like to use different colors for easy id, you can use sticks or what ever you like, just paint the top end and drive into the ground).
Once you have this visual reference system in place, you are ready to do some sample taking, mixing, and shipping.
You will be able to know which sample came from which beds and that is always nice for record keeping.

I like to use a dice (#5) pattern for building up my samples  each bed would get the five parts mixed well then take the sample from that batch, bag the leftovers for any future use or if you want to compare labs.

If you want to blanket sample a bed (this is 1 bed divided into equal parts and the #5 dice pattern is taken from each part) you can either send these in separately or blend them together prior to taking part of the total sample amount to bag for testing.
For sampling it is really good to get standardized samples I use a piece of 2" rigid conduit (what I had laying around) that has one end filed to a fairly sharp edge all the way around.
This is driven into the ground I am sampling down to a mark on the outside that is at 8 inches from the sharpened end.
To remove my sample I just use a stick made from a tree branch that fits inside the piece of pipe.
I collect the samples on a piece of heavy plastic sheeting (scrap again, even a garbage bag will work for this).
By doing this I can visually inspect the soil samples before I mix them up well and bag up a sample for my lab.

When you are marking your samples to send to the lab include anything that will help you identify which beds that sample came from so when you get the results you can know which beds it represents.
My markings for the lab go like this: Bed=Red, 10 parts (random select) or, DP (dice pattern) This way I could send in samples from the whole of our gardens and know which results go where quite easily.
I also have pages in a note book that are for each of our gardens this is for listing what and how much of each amendment I make and when I made the addition of those materials.
I have other pages that I list weather, light exposure, moisture content, bacteria present, fungal hyphae present, etc.  I do a pH test twice a year (even litmus paper works fine for this).

Redhawk

The rigid conduit idea is ingenious! You can sample even when the bed is planted. I thought a bit around and I had a bed were carrots did not grow well, so I will sample this one and one bed were carrots grew well. and I will sample another one were beets did not grow. I think three samples is expensive enough to begin with. Then I can see weather or not my beds differ very much. I can see weather there is a nutrient missing in all of them.