Choosing an Inductor for a Bass Control

[J M Fahey]

1) what happened to the original inductor?
2) if it still exists, use it; if not, junk that obsolete tone control , rewire the triode as a regular gain stage and insert a conventional, passive bass/treble control; an Ampeg type one will provide symmetrical boost/cut, flat on center, and in general do the same job but better.
Amp will be improved keeping the basic same sound signature.
It will definitely hum less.

[bob p]

1. I previously mentioned that it's gone.
2. Yes, I had already thought about going to a James/Baxandall type setup, but that recharacterizes this restoration project and changes it into a resto-mod project.

This amp was built in an era where caps were prohibitively costly and expensive. Small caps were used because big caps were expensive. Inductors were cheap. It stops being a 1930s amp if I rebuild it like an amp from the 60s.

I've given serious thought to your recommendation, but only as a backup plan.

So far I've heard inductor value recommendations that are as low as 0.1H and as high as 20H. I'm sure we'd all love to hear the results of an actual measurement if g1 can find the time to take it.

[MarkusBass]

It wasn't me who recommended 20H . My recommendation is 13H (first test it with 10H choke). It seems to me that the tone stack is cut-only. They didn't know James tone stack at that time. I wonder how does it sound. Is the amp from the 30-ties? This would mean that the amp is almost 80 years old. I also would go for the restoration of the amp. Even if it does not sound good, it may have very high value for amp collectors.

Mark

[DRH1958]

Does anyone has a photo of the amp guts?

Mark

[ATTACH=CONFIG]40120[/ATTACH]

There's two shots in this article about the EH-185 with them describing the 6J7 preamp and 6N7 PI as the schematic in Bob's OP shows and also shows the choke mounted on top by the tubes. Unfortunately in the article they mention it but no value is given but you can see the GQ 13 stamp: https://www.vintageguitar.com/21529/39-gibson-eh-185/

[bob p]

I'd seen that picture before, but I didn't realize that the transformer shown in the pic had the house part number stenciled on it. It's a pretty big inductor.


I think it's funny that they decided to use such a big inductor for a tone control. If I were installing a part that big then would have used it as a supply filter.

[MarkusBass]

I noticed that the amp "can be extremely worthy (when properly restored)". The pots in the tone stack have "Normal" in the middle position, which suggests that the tone stack is of "boost and cut" type. I have to check my simulation. The inductor is big as a choke and can have easily 13H (as the symbol suggests).

Mark

[bob p]

The schematic looks like boost and cut to me.

Was your estimate of 13H based upon the appearance of "13" in the "GC13" house number?

[MarkusBass]

Was your estimate of 13H based upon the appearance of "13" in the "GC13" house number?

No . I told you that I created SPICE simulation of the tone stack and tested it with 5H, 10H, 13H, 15H and 20H inductors. With 13H and 15H I got reasonable results. But "GQ13" symbol is also something to think of. Why didn't the named it "L1", or Q2"? I think that there was some reason for this .

Mark

[bob p]

Understood. But to be fair, I've always thought that the circuit had boost and cut, which diminished the value in my eyes of a cut-only model. I could be wrong, it won't be the first time. lol. FWIW I have also received PMs from other people whose models differed from yours by an order of magnitude, and my envelope calculations were somewhere in between. So far nobody has actually shown me their work, which makes me reluctant to accept anyone's suggestions based solely on blind faith. If anyone would be willing to post their calculations (in the form of simulation files if need be) that could allow someone else to independently examine and verify their methods.

Personally, I'm really looking forward to hearing g1's measurement results.

[Chuck H]

Some web research revealed this (below) from a Vintage Amps thread covering the same topic:

I'd seen the guts shots and that's how I come to realise the choke is what the GQ13 is on the
schematic since it doesn't say so. The tone control article by Thordarson I found after I posted here. It only identifies the "tone control choke" by it's part number T-14C70 not the value. So I looked up an old Thordarson catalog from 1938 here: http://www.bunkerofdoom.com/xfm/THORDAR ... on400C.pdf and low and behold it shows that part as a "tone control, hum bucking type" interestingly enough and the specs are 22H, 230R, 5mA max.

If this is indeed the animal then I'd say the humbucking winding is going to null any aftermarket replacement possibilities. And they probably wouldn't have made it that way unless it was necessary. I'm now siding with others that are suggesting rewiring for a more conventional TS.

[g1]

The Thordarson is what is referenced in some article about chokes in tone controls. It is not the unit in the Gibson amps.
Vintage Amps Bulletin Board ? View topic - Gibson EH185

[J M Fahey]

1) yes, itīs cut/boost.
It cuts when shunting highs or lows to ground from 6J7 plate, boosts when grounding 6J7 cathode.

2) it is not symmetrical, neither in amount of boost or cut nor, much more important, *frequency* : when boosting C and L are bypassing 10k ohms; when cutting they re bypassing whatever impedance is on 6J7 plate, which is not plain 100k since they are in parallel with 6J7 Rp, ots internal plate resistance.
Since itīs connected as a triode, itīs relatively low.

Donīt know its value but a similar 12AX7 triode stage has 100k in parallel with internal 68k (read it in some datasheet) so effective plate impedance is around 40k.
Donīt think we have a way too different value there, but in any case itīs easy to measure.

3) There is also another "invisible" resistor there, one in series with 6J7 cathode.
Again can be calcuated but measuring is easy and accurate for that tube, instead of a generic one.

Momentarily disconnect the tone control by lifting one end of the .5 cap and the wire going to .1-6N7 grid.
Momentarily add a 500k resistor from grid to ground so it does not lose bias.

Then inject 100mV 440hz tone in 6J7 grid, measure audio on its plate, I guess youīll have around 5X gain.

* Measure max gain/boost (and internal cathode resistance):

Ground .5uF cap , what does plate signal rise to?

Say it rises 4X , then max boost is 4X=12dB
And internal resistance is 10k/4=2500 ohms.

4) Now you can also trace tone control curve and turnover frequency:
It will start rising above 10k-.01uF or 1600 Hz, a very reasonable value.
It will probably shelve at some 4x that, or 6 kHz, again very reasonable for a guitar.

The inductor "should" probably boost below, say, 800 or 1000Hz to usefully match the Treble control range, so calculate what inductance value shows some 10k impedance at 800 or 1000Hz and there you have it.
I bet itīs quite a hairy inductor.

Iīm analyzing this from a purely functional point of view and quite certain that the original designer must have travelled a similar path.

5) * Now analyze Cut:
it will work about the same, but it will be stronger (cutting is easy, boosting not so) and will happen at a different frequency.

Not sure about this, it depends on personal taste, way back then people were used to cutting highs, what simple tone controls do, but must have cherished their hard to get lows (poor speakers and low power) so I donīt see them cutting them off.

Notice that Fender tone controls which became way more popular were basically boost only.

6) another path is to compare it to old Gibson amps, some used a Cut-only Bass control (Matchless amps copied that) , wouldn`t be surprised at all that said passive control turnover frequency was similar to what was used here.

Designers usually have some favourite "pet ideas" and tend to repeat them all over the place, we might call them their "signature".

[MarkusBass]

OK, I have to admit that I was wrong. The circuit is cut/boost. The reason I was saying that it is only "cut" was that I did not use high enough value of the inductor and for TREBLE pot I was getting +3dB boost only. The same for the BASS pot. But once I changed the inductance value to 22H, everything looks almost perfect. Just take a look at TREBLE and BASS pot frequency response. I think it looks exactly as it should. I would just find, or wind 22H inductor. The reason why it is big but not huge is that the current through it is in range of uA or even less.

Mark

[nickb]

Here's a different approach that will perform similarly, be more flexible and cost about $2... a FET based gyrator instead of the inductor.


Here is the response with a 22H inductor:



and with the gyrator:



To change the inductance vary R2

[g1]

But once I changed the inductance value to 22H, everything looks almost perfect. Just take a look at TREBLE and BASS pot frequency response. I think it looks exactly as it should.

I think we have a winner.
I had a chance to do some quick checks although I did not have a chance to disconnect the inductor as it's fairly involved and was just visiting where the unit is.
Measured with DE-5000, inductance was 21.8H, Q of .86, DC resistance about 2.5K.