MPP preamp is vastly improvable

[waltr]

Hi 6666
I finally was able to make a little progress with this at least the first design on post #3.
I did not do the integrator changes at least not yet.

I built both a MPP rev e preamp as a stand alone and the three amp design for comparison.
Using the Mpp layout I added the third amp and used Teleno's recommendations for amps.

Here is the three amp set up.
I'm just including the 4562 in the tests from here on out as there is no reason to move forward with the other two.

[ATTACH]49987[/ATTACH]
Thanks Again Teleno

Can you please post a schematic of what you built. I went back through this thread and was unsure of exactly what schematic is correct.

Looks very, very good.

[godigit1]

Is the preamp test saturation/clipping recovery or small signal gain with no clipping?

Hi bklein,

The test I did was just signal time to 0v .

I input a 50k square wave at 2mv.

Ive gone back and looked at the slew a bit .

With the stock Mpp pre amp and a 4562 im seeing 5.12 mhz before the first sign of slewing.
With the Tlc081 as a third amp Im seeing first sign at 2.122 mhz.
Looks like the third amp is limiting bandwidth .

However using a square wave looking at the time where the decay slope at turn off meets the turn on signal is 31k faster with the TLC081 set up.
Stock Mpp With 4562 meets at 200k and 231k for the three amp set up.

Hope that make since,

[godigit1]

Hi waltr,

Here is the schematic I built to, There are three extra spots at the feedback resistors I was going to use for further testing.

I also made up a quick pcb layout for it.
It has no ground plane or outline pretty rough I made it for easy perf board layout.

[godigit1]

Hi Godigit1, wow you have been busy, and you have done the testing I was planning to do, when I got the flu and 3 colds last year my projects got put on hold, lately I have been working on 5 volt micro controlled PI circuits just for fun, my tlc081 arrived but I think my 4562's are fake, but it looks like your testing has proved the preamp a winner, thanks.

Hi 6666,

Bummer about the 4562's I'm curious what are they doing or not doing that makes them bad?

[waltr]

Thanks.
So this is without the JFET Clamp.

[godigit1]

Hi waltr,

I think Q3 is intended as first sample, Teleno had included the first integrator change at this point.

I built this version but with no luck.
I've since found one of my mistakes a mistake I keep making on Jfets.

[6666]

Hi 6666,

Bummer about the 4562's I'm curious what are they doing or not doing that makes them bad?

Bummer yes, they give the same wave form as 5532

[6666]

Hi waltr,

I think Q3 is intended as first sample, Teleno had included the first integrator change at this point.

I built this version but with no luck.
I've since found one of my mistakes a mistake I keep making on Jfets.

I hate jfets what was your mistake ?

[godigit1]

I hate jfets what was your mistake ?

A couple times now I have made pcbs and going too fast wired pins to schematic and not looked at gate sources drain actual locations.

Sounds like their just relabeling 5532's , Hard to believe its worth their while.

Best of luck to you next time.

[godigit1]

I hooked up the three amp set up to the Mpp.
I was a bit surprised to see that its a bit slower than the Mpp dual amp with just the 4562.
It looked really good on the signal generator.

This is what I'm seeing on the stock MPP preamp with a 4562.



This is the 3 stage amp.



Here is the 4562 before the Tlc081, the 081 slows thing down quite a bit as you can see in the above picture.



Here is the Tx fly back.




As far as performance the first big difference is the damping resistor value.

Using the same coil for both amps, the stock Mpp damped at 520 ohms and with the 3 stage its damping at 1.637k
Seems to have picked up a couple inches but the threshold is a bit unruly, that may be partially my test set up.
In the pic Below you can see its off board and wires sticking up. I think I'm picking up more environmental noise . plus no ground plane on the pcb.



I tried a Tlc082 and 081 combination with pretty much the same results.
I think on board dead bug style may be next then another half of a 4562 for the third amp, otherwise the stock amp is already faster then the sample generator. Mine shuts down at right under 5us for first sample spacing.
I would like to get another 1us or so, that way I have a little leeway on a 5 us sample.

[6666]

the stock Mpp damped at 520 ohms and with the 3 stage its damping at 1.637k

You are getting some varying results, but you can see a difference, the damping changing is a little curious.

[waltr]

Thanks for posting O'scope pictures.

I take it that your signal generator did not Saturate the first op-amp like the TX coil's Flyback does.
It is this flyback saturating the op-amp that is the issue.
1- many op-amps take some time to 'recover'.
2- Due to Gain the op-amp output is only a Very small fraction of the fly-back amplitude. so hard to compare on a scope.

Hint: Does your scope have an External Trigger input?
If so then feed the TX pulse logic signal into the external input and select this to trigger scope.
Now you have two scope probe inputs so can look at TX fly-back AND Pre-amp output. Or at two points in the pre-amp stages.
There is no need to 'see' the logic pulse signal since you know what this looks like.

Good stuff.

[dbanner]

You are getting some varying results, but you can see a difference, the damping changing is a little curious.

That is indeed curious. That's a big difference in the damping value for the same coil.

I would have thought that the coil only sees the first stage, and the capacitative difference would have been small between the stock circuit and the 3 stage circuit.

[green]

That is indeed curious. That's a big difference in the damping value for the same coil.

I would have thought that the coil only sees the first stage, and the capacitative difference would have been small between the stock circuit and the 3 stage circuit.

Read post #1. Circuit Rd=damping resistor if amplifier doesn't keep one end of Rin at ground. Circuit Rd=(damping resistor*input resistor)/(damping resistor + input resistor)if amplifier keeps one end of Rin at ground. Rd changes depending on what amplifier does.

[dbanner]

Ok. After reading the post, it becomes clearer now. Interesting.

Dealing with transients in the uS range is a bit daunting.
That's where all those electrical characteristics you see on those opamp datasheets really come to the fore.