THAT 202XL recreation

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
ilya
Posts: 67
Joined: Tue Dec 29, 2009 3:10 pm
Location: Russia
Contact:

Re: THAT 202XL recreation

Post by ilya » Fri Nov 24, 2017 12:26 pm

mediatechnology wrote:
Fri Nov 24, 2017 11:21 am
I don't think that's correct.
If you look at the 2181 datasheet page 2 table 3 it shows a +0.33% gain control tempco.
You still need -0.33% to correct it.

Also see equation 1 on page 5 of the 2181 datasheet for the delta-T term in the denominator: http://www.thatcorp.com/datashts/THAT_2 ... asheet.pdf
Indeed. That's really odd, I mean, why an THAT application engineer would provide completely incorrect info?... Gonna send them another mail.

User avatar
JR.
Posts: 2418
Joined: Sat May 24, 2008 7:21 pm

Re: THAT 202XL recreation

Post by JR. » Fri Nov 24, 2017 12:53 pm

enjoy the math (I don't) http://www.thatcorp.com/datashts/dn128.pdf

JR

ilya
Posts: 67
Joined: Tue Dec 29, 2009 3:10 pm
Location: Russia
Contact:

Re: THAT 202XL recreation

Post by ilya » Fri Nov 24, 2017 1:18 pm

JR. wrote:
Fri Nov 24, 2017 12:53 pm
enjoy the math (I don't) http://www.thatcorp.com/datashts/dn128.pdf

JR
Ugh..... Seen than. This stuff is far above my math grade unfortunately((((

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: THAT 202XL recreation

Post by mediatechnology » Fri Nov 24, 2017 2:56 pm

One thing slightly off-topic is that the distortion in the pre-trimmed 2180 has a lower tempco (of distortion change) than the externally-trimmed 2181.

The 2181, of an equivalent grade, can be trimmed lower than the 2180 but the 2180 will out-perform it over temperature. If I were going to do a "202" I think I'd do it with 2180s.

I still think you need a -3300 ppm (NTC) resistor to correct the scale factor tempco or a PTC working as an NTC.

montemcguire
Posts: 4
Joined: Fri Mar 30, 2018 9:26 pm
Location: Watertown, MA USA

Re: THAT 202XL recreation

Post by montemcguire » Sat Mar 31, 2018 10:33 am

I'm trying to use a THAT RMS detector for a project, and also discovered that the Panasonic 3300ppm/K resistors are now EOL and unobtainable. The solution I settled on was to use a Platinum 500Ω RTD along with a conventional resistor to adjust the Pt tempco of 3850 down to 3300. The idea is that if you place a conventional essentially zero TC resistor in series with a 3850ppm/K Platinum resistor, you can dilute the tempco of a 500Ω Pt resistor down to 3300 ppm/K:

3850/3300 = Rx / 500Ω, where Rx is the total of the 500Ω Pt RTD and the padding resistor. --> Rpad = 83.3Ω

Vishay Beyschlag makes the PTS series, and the PTS080501B500RP seemed to be a good candidate. My world is SMD anyway, so these parts are preferable to me. The other advantage of the Platinum parts is that the tempco and resistance tolerance is a lot tighter than the old Panasonic parts, probably removing the need for random trimming. These parts are not all that cheap, but there doesn't seem to be any good way around this problem other than using a goofy tempco resistor.

ilya
Posts: 67
Joined: Tue Dec 29, 2009 3:10 pm
Location: Russia
Contact:

Re: THAT 202XL recreation

Post by ilya » Sat Mar 31, 2018 11:22 am

Thanks for your input.
It's a pity I haven't seen those platinum tempcos before I put an order with Mouser. What I decided to try was Vishay TFPTL series (https://www.mouser.com/ProductDetail/71 ... 0L1000FH5B). They have 4100 ppm +/-400 tempco though, so your SMD platinums look like a more suitable and precise option. I need a 100R shunt to gnd, so a padding resistor with your temcos works out to something like 17R according to your formula (or 24R for TFPTL parts). Going to play with a hairdryer and see how the circuit behave...

ilya
Posts: 67
Joined: Tue Dec 29, 2009 3:10 pm
Location: Russia
Contact:

Re: THAT 202XL recreation

Post by ilya » Sun Jun 24, 2018 2:52 pm

A quick update.
The 2002T curcuit is working fine with diluted tempco as suggested by montemcguir. Thanks for sharing the idea.

I'm having a hard time with "VA ADJUST" trimmer though. As is shown on the schematic, the total resistance can be varied between 10k and 60k. As far as I know, this control adjust TDH at high gain reduction levels. Well as it is in the original schematic, it doesn't make any significan difference when I adjust the trimmer. However, if I short out the 10k resistor, I can achieve sognificantly lower THD. I've measured the resulting resistance and got around 1.5k. Any ideas if this is the correct behaviour?

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: THAT 202XL recreation

Post by mediatechnology » Sun Jun 24, 2018 3:07 pm

ilya wrote:
Sun Jun 24, 2018 2:52 pm
A quick update.
The 2002T curcuit is working fine with diluted tempco as suggested by montemcguir. Thanks for sharing the idea.

I'm having a hard time with "VA ADJUST" trimmer though. As is shown on the schematic, the total resistance can be varied between 10k and 60k. As far as I know, this control adjust TDH at high gain reduction levels. Well as it is in the original schematic, it doesn't make any significan difference when I adjust the trimmer. However, if I short out the 10k resistor, I can achieve sognificantly lower THD. I've measured the resulting resistance and got around 1.5k. Any ideas if this is the correct behaviour?
Glad you got it working with available tempco resistors.

I don't think VR2 is the actual symmetry trim. I don't remember exactly but it looks like the actual trim is external at header 1 pin 7 labeled SYM. I think VR2 "trims the trim."

I think we need to look at the actual 2002T datasheet to see what trim network connects to pin 7 of the 2002T.

I'll search here on the forum to see if I have a link to the 2002 and on my local drive.

EDIT: Have a look here at the various trim circuits: viewtopic.php?f=12&t=34&p=194

I would guess from the looks of it that the open end of R7 would connect to the wiper of a 50 K trim pot tied to +/-15V at each end.
That trim is the actual SYM THD trim.
You want to do that with Ec at 0V. (And I think VR2 at maximum.)

VR2, and similar methods on the other retrofit VCAs, appears to inject an Ec-dependent correction to alter SYM vs Ec.
My hunch/uneducated guess is that VR2 extends the dynamic range/THD performance at one extreme of Ec most likely gain.
But I'll caution you by saying I don't know.

EDIT 2: VR2 could also be used to minimize control voltage feed-through.

I would install an external trim, adjust minimum THD at 0 dBu level and at 0dB gain (Ec=0v).
Then increase gain, see if you observe a rise in THD, and lower VR2 to see what difference it makes.

I'm darn sure that VR2 is not the actual THD trim though.

ilya
Posts: 67
Joined: Tue Dec 29, 2009 3:10 pm
Location: Russia
Contact:

Re: THAT 202XL recreation

Post by ilya » Sun Jun 24, 2018 3:41 pm

My apologies for not being precise and for possible confusion. Of course there should be an external trim circuit for minimizing THD. As you said, this is often a voltage divider connected between +/- rails and fed into the sym pin of the VCA. I have this arrangement, I just didn't mention this fact. What VA ADJUST is supposed to do is lower the THD at high gain reduction levels (as far as I know). My current trimming procedure is as follows: apply 0V to EC pin and trim for minimum THD, apply voltage to EC pin to get 15-20 dB of gain reduction and trim the VA trimmer for lower THD. And it is at this point that I find that I need drastically lower resistance of VA trimmer in order to get lower THD.

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: THAT 202XL recreation

Post by mediatechnology » Sun Jun 24, 2018 3:52 pm

Thanks for clearing that up.

The more I look at this I think it is an anti-feed through adjustment.
I went back and looked at a couple of old emails from Gary Hebert regarding the 202/2002 retrofits.

This original 4 VCA design did not use differential Ec control.
The client it was designed for needed attenuation of -110dB at mute and it would thump at such deep attenuation due to gain cell transistor saturation.
I've seen that as well and have avoided it using differential Ec.
This trim may be to reduce that.

If you look at the 2181 datasheet they make the comment that the optimum THD point and point of minimum feed through are usually the same, or close, but not always so.
I think this trim may be for the not always so situation by injecting Ec into SYM to reduce control feedthrough.

The later VCAs used 8 2181 and differential Ec control.
On one of them I looked at 7 of the VCAs did not have SYM adjustment only 1.

One thing that bothers me in the original schematic is the NJM2114 follower with no feedback current limiting through the input diodes during start-up. That is a latch-up opportunity for a 2114, 5532 or 5534.

Edit: I also remember a comment from either Les or Gary about the futility of THD trim at any Ec other than 0V because its trimming out other stuff not related to internal Vbe mis-match. (Which is what the trim fixes.)

So I'm pretty sure VR2 relates to Ec feed-through.

Post Reply