T-Pad behaviour

[Che_Guitarra]

I'm trying to isolate a problem in a 1176 build.  Circuit testing seems to indicate I have a problem either with the input transformer or T-pad input controller.

I've never dealt with a T-pad before, so i'm not sure what is considered normal behaviour.  But with a DMM in hand it seems this item has an unusual response to measurement.  I'm not sure how to quantify it... but pass signal through it and you could perhaps describe it as a concave response.

Is this normal?  What is typical T-pad behaviour?

Maybe a longshot, but are there any graphs to demonstrate what I should seeing?

 

[pucho812]

are you sure it's wired correctly? I would double check your wiring to and from the t-pad

 

[ruffrecords]

Can you be more specific about what you mean by circuit testing?

Cheers

Ian

 

[Che_Guitarra]

I'm having a problem with the off-pcb part of the input circuit in my Rev D 1176 clone.  It's a component failure of either the t-pad or input trafo.  I can confirm this for certain as in the last 24 hours my Rev A 1176 kit arrived in the mail - the first thing I did was build up the input circuit and patched it into my Rev D's PCB, and with this input chain I can get my Rev D working perfectly.  I've also re-wired it's input chain to eliminate the possibility it was a wiring error on my behalf.


So I can point to either the t-pad or the input trafo, I want to know what typical patterns I should see of the t-pad with my DMM when tested with a probe?

 

[abbey road d enfer]

I want to know what typical patterns I should see of the t-pad with my DMM when tested with a probe?

In a textbook T-pad, you would see constant input Z and constant output Z, and as a consequence input-to-output resistance would be constant.
Nevertheless, physical T-pads are not perfect and you may see some variation.
IMO, a DMM is not the proper instrument to investigate a T-pad. The best way is to inject signal in the input, measure the output and draw the Attn vs. rotation graph. In any case, fully CW should produce none or very little attenuation.

 

[Echo North]

The t-pad tries to maintain a 600 Z.  I forget the exact measurement, but it will vary by maybe 100 Z, usually low.

 

[Echo North]

IMO, a DMM is not the proper instrument to investigate a T-pad. The best way is to inject signal in the input, measure the output and draw the Attn vs. rotation graph. In any case, fully CW should produce none or very little attenuation.

Yes exactly, try this.

 

[Echo North]

ohm out the input transformer, then it's off your list if ok.

I did that with him in the 1176 thread.  The input tx looks fine IMO.  Winding DCR is good, no shorts.

 

[Che_Guitarra]

Seeing I can't 100% validate either component i've decided to replace both... no doubt if I only order a replacement t-pad the issue will end up being with the input trafo, meaning another two week wait and another bout of postage to pay.

Oh well.  The order is in and a guaranteed solution is on it's way.

 

[emrr]

I'm looking at one from CAPI, same as the Hairball pieces.  Says right on it, 2 outer decks are 500R, center is 1000R. 

Measure from outside legs of each deck, should be in those ranges per deck. 

Wired properly, turned all the way down, the outer decks will measure about 500R ea and the inner deck will be fully shorted, a fraction of an ohm.  In to Out will measure about 1000R, In or Out to Common will measure about 500R. 

A proper vintage T will measure more like 5K on the center deck and some will go open when fully up, effectively disconnecting itself entirely.  These new ones have slight insertion loss, but are just fine for most things. 

 

[Echo North]

Purple originally tried a 5K pot in the center (still shown on the MC77 schematic) but the performance was off for some reason.  They ended up getting the best performance with 1K center pot, in the 1176 anyway.

Mike

 

[Echo North]

If you confirm one of the components is crappy, email me and I'll send you a refund.

Mike

 

[emrr]

I'm sure Purple couldn't get the required taper in a 5K deck.

I was just putting 2 of the modern ones onto the support PCB, and was reminded of how much I hate that PCB.  I assume CAPI and Hairball share the PCB design.  13 legs to align with tiny form fitting holes; it's really tedious, and I always end up bending a leg, possible to break one off.  I would redesign the PCB to have larger holes for sure.  No reason not to do so, once soldered in place the legs care not how large the through hole is. 

I'd really rather see these with solder lugs, and no need for a PCB.  That would be far easier to assemble.   

 

[Echo North]

I'm sure Purple couldn't get the required taper in a 5K deck.

I was just putting 2 of the modern ones onto the support PCB, and was reminded of how much I hate that PCB.  I assume CAPI and Hairball share the PCB design.  13 legs to align with tiny form fitting holes; it's really tedious, and I always end up bending a leg, possible to break one off.  I would redesign the PCB to have larger holes for sure.  No reason not to do so, once soldered in place the legs care not how large the through hole is. 

I'd really rather see these with solder lugs, and no need for a PCB.  That would be far easier to assemble. 

Good point, the PCB mounting is tedious for sure.  At the very least I can expand mounting holes.

 

[emrr]

Thanks for listening, figured the FB was worth throwing out there.  Before you guys were doing these, there was a group buy on T's with solder lugs, may have been a different manufacturer.  Anyone know a reason they can't be had anymore? 

 

[Echo North]

Thanks for listening, figured the FB was worth throwing out there.  Before you guys were doing these, there was a group buy on T's with solder lugs, may have been a different manufacturer.  Anyone know a reason they can't be had anymore?

Yes the old 71(?) series clarostat.  Jeff and I sold those for a long time.  They kept wanting us to buy more and more and kept raising the price on both of us.  I think in the end they wanted a min 2K pcs order with no savings.

I think that series is being discontinued.

Mike