Brownface Vibrato in the Gallery

[R.G.]

1. i have a quite serious volume drop compared to the clean signal. do you think this is becasue i used 5458's instead of 5952's..

Could be. The 2N5458 has a lower gain than the 5292.  Could also be a construction bug. DC voltages?

2. how does the wiring to the input ring affect normal power bypassing.why does it have to pass through the input jack? can it be routed straight to ground? i usually use the ring as a power cut to ground so the pedal has no power with no cable jack inserted. i usually wire the board grounds to the input jack ring and the power ground to the input sleeve..

The wiring to the input ring is a superior method to replace wiring the battery negative to the input ring. If you wire the input ring to the battery negative, all of the pedal's current runs down that wire, and so the input wire to the effect which also carries the input signal has the battery current on it. This makes a sewer ground out of your most sensitive signal refererence ground. While you get away with it a lot, it's a good way to get switching noise and oscillation. So I just don't do it any more.

Instead, I wire the battery and DC jack directly to the circuit, but the positive side goes through a PNP transistor. The base of the PNP is pulled down by the input jack ring, and this causes the PNP to turn on when a plug is inserted. It's the same action as wiring battery negative to the ring, but now the current in the input signal ground is only the base current of the PNP. It's both much smaller than the entire pedal current and also constant. This means that there is no chance for input noise from it.

The PNP circuit does what the ring circuit used to do, and does it better. It's a replacement.

replacements made:
5458's instead of 5952's

MIght be your gain problem. Maybe.

220pf instead of 250pf
.047uf instead of .05uf
.0047uf instead of .005uf
used 1uf tant caps for the non-polar 1uf caps.

No particular consequence.

[markm]

I used 2n5952's in my build, I don't have a volume drop issue so, it may very well be due to the 5485's.
Please note the 2 matched pairs in the layout.
They don't need to match each other but, it is a good idea to match the 2 pairs as indicated.

[markm]

Oh man!  hadn't been on here in a while.  I've been looking for a pedal to do this particular vibe for years now.  I was fortunate enough to play a brownface vibrato and was totally hooked on the sound.

I'm completely pumped about this, just need to order some of the trannies now.  Mark, thankyou!

Check out the sound samples of this circuit from Ulysses.
Great sounding Brownface Vibrato!
I'm amazed at how "vintage"  it really sounds. I mean,....It sounds "Old"! Very Cool!! 

[ulysses]

hey guys

in an effort to try and debug my build i tried J210's in the input and output buffers. didnt make any difference.

i did a trace on the schem against the layout and found the following errors
1. c9 - incorrect orientation on marks layout
2. c10 - is supposed to be polarised and it is not on marks layout.

i put c9 in correct orientation and it didnt make any difference.
i put my tant cap for c10 which has polarity markings in the correct orientation and it didnt make any difference.

i went over all solder joints and they look fine - no trace bleeds

one thing i did notice was that betwwen the low end and high end of the sweep the volume does have a percieved loss of volume due to the bass cut - but im sure there is an isolated volume loss somewhere..

i did the voltages on the trannies - if you guys could compare i would appreciate it

Q1:
D 8.93
S 3.73
G 3.16

Q2:
D 5.62
S 2.19
G 1.47

Q3:
D 5.69
S 2.15
G 1.41

Q4:
D 8.93
S 3.87
G 3.38

Q5:
D 1.19
S 1.19
G voltage speed change varies on the speed of the dual speed pot (high 1.2 / low 0.01)

Q6:
D 1.19
S 1.19
G voltage speed change varies on the speed of the dual speed pot - (high 1.3 / low 0.01)

Q7:
C couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage
B jumped around - high 0.7v
E jumped around - high 0.08v

Q8:
C 8.94
B couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage
E couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage

Q9:
C 6.67 (with speed pot at 0%) (kept jumping around 5-7v with speed up)
B 2.8
E 2.2

Q10:
C 9.0
B 8.25
E 8.94

cheers
ulysses

edit: one other strange thing that i checked before soldering in the dual pot was - the top half of the pot had a resistance of 50k, the bottom half had a resistance of 55k.. i doubt this would affect the output volume though..

[markm]

Well, I wouldn't think C10 would make a difference as Polys are usually considered to be better than Tants and electros for "tone".
C9, ummmm, that's an oversight on my part but, Hell.......Mine Works! No Volume drop.
Perhaps you should go with the 2N5292's?
One thing I discussed with R.G. about this circuit was the fact that I wanted to build it correctly and not cut corners or put a "band-aid" on anything.
Which I think was the best approach to take as the man REALLY knows his stuff! He helped me perfect it and I think the suggested values should be pretty much adhered to as much as possible for complete satisfaction. Once one starts to go astray and begins to sub parts,
it becomes a whole new can of worms.
R.G. has also stated that when deviating from the specified devices, bias issues will more than likely occur which is why I went with all of the suggested parts in the layout.
If you'll notice too, the Speed Pot I chose was a B50K dual as with the 100, the first half of the throw was for the most part, useless in the circuit. The resistors to the speed control of the LFO are not 1K but rather 2.2K to deal with the stalling issue the LFO had on my first build.
The schem also shows the 220K from the .05uF off of Q1 attaching to ground and then to the junction of another 220K, a 250pF cap, and a 1M/4.7M junction at the gate of Q3. Just to let ya know, this is incorrect in the schem so, If you have changed this as well, that could very well be the problem.
R.G. has informed me via e-mail that the schem will be updated at his site.
My conclusion is ..........if this circuit is built, exactly as the Layout here shows,

http://aronnelson.com/gallery/album76/ProVibrato_LAYOUT_revised_2

It will function correctly, it will sound Very, Very Good, and you will be one happy camper!
Once the schem is updated, I'm sure some of the errors will then be understood.
I know all of this because, I built this circuit 3 times BEFORE it was posted. 

[R.G.]

i did a trace on the schem against the layout and found the following errors
1. c9 - incorrect orientation on marks layout
2. c10 - is supposed to be polarised and it is not on marks layout.

I have some updates to the schem to do, as Mark said.

one thing i did notice was that betwwen the low end and high end of the sweep the volume does have a percieved loss of volume due to the bass cut - but im sure there is an isolated volume loss somewhere..

That seems to suggest to me that you're not getting any treble signal.

Q1:
D 8.93
S 3.73
G 3.16

Looks OK.

Q2:
D 5.62
S 2.19
G 1.47

Looks OK.

Q3:
D 5.69
S 2.15
G 1.41

Looks OK.

Q4:
D 8.93
S 3.87
G 3.38

Looks OK.

Q5:
D 1.19
S 1.19
G voltage speed change varies on the speed of the dual speed pot (high 1.2 / low 0.01)

Q6:
D 1.19
S 1.19
G voltage speed change varies on the speed of the dual speed pot - (high 1.3 / low 0.01)

Looks OK.

Q7:
C couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage
B jumped around - high 0.7v
E jumped around - high 0.08v

Q8:
C 8.94
B couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage
E couldnt get an accurate reading - fluke kept jumping and saying dangerous voltage

Looks OK, LOF is working. I don't know why your Fluke is saying "dangerous" unless it just thinks
AC voltages are dangerous.

Q9:
C 6.67 (with speed pot at 0%) (kept jumping around 5-7v with speed up)
B 2.8
E 2.2

Looks OK.

Q10:
C 9.0
B 8.25
E 8.94

Looks OK.

Overall, the DC setup is OK, there's no reason there for what you're describing. It is possible that there's a wrong value part or a bad connection in the frequency separating section of the circuit, after Q1 source and on the way to Q2 and Q3 gate. It is also possible that Q5 and Q6 are very different from one another.

I told Mark that Q2 and Q3 should be reasonably matched, and also Q5 and Q6. Q2 and Q3 don't have to match Q5 and Q6, just two modestly matched pairs. This is for two reasons. First, Q2 and Q3 need to match moderately so they will have about the same amount of gain. Otherwise, the sound will always be bass-heavy or treble-heavy, depending on which one has the most gain. Likewise Q5 and Q6 need to be similar or you'll get less variation on the bass side or treble side. Matched is probably too strong a word. Very similar is probably better, as you don't need the same matching as you have in a phaser.

So take a look at that stuff. Maybe try swapping Q5 and Q6 and seeing if the change follows the transistors. If it does, match a pair and sub them in.

[R.G.]

I updated the schemo at GEO.

By the way, I did a layout of the circuit, and it can be fit into 3.25" by 1.8". To do this you need accurately sized footprints and probably a real layout program. The now-freeware Autotrax from Protel/Altium should do it.

[markm]

R.G.,
Do you have a link for this program?

[R.G.]

Sure. Go to http://www.altium.com/Community/Support/Downloads/ and scroll down to the freeware downloads.

If you can find a used copy of Traxxmaker (or Circuitmaker, the suite product) that's even easier to use. It was never freeware, unfortunately, but the company making it has gone under and you can sometimes find a used copy.

The trial version of Traxmaker is here: http://www.elektroda.net/download/file586.html but it's a limited version. I don't know if you can even buy the upgrade code to make it into a full version, since Microcode Engineering is out of business, I think.

[ulysses]

thanks for the replies guys.

i dont think i am getting a treble loss. just an overall volume drop. if you listen to the demo i posted in my section of the gallery/builds/vibra-tone you can see how it sounds. it does sound very good, just lower volume than the dry signal.

i matched the pairs exactly to 3 decimal places. 2x 0.901volts & 2x 0.912 volts - would an exact match have an effect on volume?

i checked all the caps last night. i will check every resistor later tonight when i get time.

i was aware that you modded the speed pots and attached resistors - i left them as per the layout.

thanks again to mark and rg for their work on this one

cheers
ulysses

[R.G.]

i matched the pairs exactly to 3 decimal places. 2x 0.901volts & 2x 0.912 volts - would an exact match have an effect on volume?

That's certainly sufficient, but probably not necessary. No, having an exact match will not drop volume. It's possible you're just getting low gain out of Q2 and Q3.

If you don't find a wrong component value, you may need to change Q4 to a gain stage. You would do this by changing R14 to 4.7M, inserting a 3.3K resistor between Q4's drain and +9V, and taking the output signal off the drain by connecting C10 to the drain instead of the source. If that's still not enough, bypass R16 with a 10uF cap.

But I really think there's something else wrong.

[ulysses]

hey guys

just an update

i checked all the resistors. they were all correct.

i experimented with the polarity of the tant 1uf caps. i found that in certain orientation the lfo would die after about 1 minute. so i put them back to their original orientation.

i swapped all 5458's for 5485's matched almost perfectly. worked exactly the same with no volume increase.

i will order some 5952's and swap them in the cradles. hope i will get a volume increase

i have been over it heaps of times and im sure i've got it correct.

will let you know when i get the 5952's if there is a volume increase. mark, you dont happen to have a bunch of 5458's or 5485's you could test for me do you? - to see if there is a volume drop?

cheers
ulysses

[markm]

mark, you dont happen to have a bunch of 5458's or 5485's you could test for me do you? - to see if there is a volume drop?

I think you may have a different issue here.
I tested all 2N5485's in my last build and it was fine......
Can't say as to which way you should go with this?
I'm going to be outting together my final build of this circuit this week as I am on vacation and plan on starting to box it up as well.
Maybe R.G. has some ideas as to what else you can do?
Your audio samples sound great though ulysses, real nice indeed. 

[ulysses]

hmmm...

rg, would 16v polar caps be the cause of the problem? should i bump them all up to 63volt ones?

cheers
ulysses

[markm]

I don't think this should make any difference at all.
16V vs. 25 vs. 63 should not matter. I did use Polys for the LFO section as they are much more tolerant in that application.
My build has 16V electros for the polarized caps. You don't suppose one of the Tants gave up the ghost in there do you?

**EDIT**
I have to ask but, is the 9V your using Fresh?
I did notice that when my battery began to crap-out, volume became a problem......just a thought.

[ulysses]

hey mark

thanks again for the reply

im using a boss psu to power it. so it is fresh

i was just wondering about the caps in case anywhere in the circuit boosts the voltage up past 16v.. clutching at straws..

do you know how to match Q5 & Q6 to 100uA? i have them voltage matched using the jfet matcher on geo.. but im not sure how to match them to 100uA like it says in rg's updated pdf.

cheers
ulysses

[markm]

do you know how to match Q5 & Q6 to 100uA? i have them voltage matched using the jfet matcher on geo.. but im not sure how to match them to 100uA like it says in rg's updated pdf.

I don't.
I just matched them as they were going to be used for a phase shifter using the GEOFEX matcher. That's it.
Luckily, out of 10 I had 2 really good matches and those are what I used.

[audioguy]

The wiring to the input ring is a superior method to replace wiring the battery negative to the input ring. If you wire the input ring to the battery negative, all of the pedal's current runs down that wire, and so the input wire to the effect which also carries the input signal has the battery current on it. This makes a sewer ground out of your most sensitive signal refererence ground. While you get away with it a lot, it's a good way to get switching noise and oscillation. So I just don't do it any more.

Instead, I wire the battery and DC jack directly to the circuit, but the positive side goes through a PNP transistor. The base of the PNP is pulled down by the input jack ring, and this causes the PNP to turn on when a plug is inserted. It's the same action as wiring battery negative to the ring, but now the current in the input signal ground is only the base current of the PNP. It's both much smaller than the entire pedal current and also constant. This means that there is no chance for input noise from it.

The PNP circuit does what the ring circuit used to do, and does it better. It's a replacement.

This method looks and sounds really interesting.... is there more info posted on it anywhere?

Thanks!

Audioguy

[ulysses]

If you don't find a wrong component value, you may need to change Q4 to a gain stage. You would do this by changing R14 to 4.7M, inserting a 3.3K resistor between Q4's drain and +9V, and taking the output signal off the drain by connecting C10 to the drain instead of the source. If that's still not enough, bypass R16 with a 10uF cap.

But I really think there's something else wrong.

hey rg

i made all the changes you suggested. i didnt have a spair 10uf cap so i used a 100uf cap instead.

first i made the changes to R14, added the 3.3k and changed the output signal path. there was a volume increase, but still not unity.

i then swapped r16 for a 100uf cap. i didnt get any output at all. i changed the polarity of the 100uf cap to "+ to ground" and it sparked right up. it now has a little higher output than unity. but it is much closer to unity than before. you might say it has a tiny ammount of boost. which is prob a good thing for this effect.

just wanted to say thanks for taking the time to tell me how i cound resolve my issue. i appreciate it.

cheers
ulysses

[R.G.]

just wanted to say thanks for taking the time to tell me how i cound resolve my issue. i appreciate it.

My pleasure. How's it sound?

The PNP circuit does what the ring circuit used to do, and does it better. It's a replacement.

This method looks and sounds really interesting.... is there more info posted on it anywhere?

Like most of my "secrets" I've hidden it away - at GEO. See http://geofex.com/Article_Folders/wrapper/wrapper.htm down around "beyond the third generation" from 2002, and before that in the Hum-Free Splitter articles in 1999. One of those has a three transistor circuit that turns on both + and - side batteries from only the connection to ring in the input jack.

And in any case - what more information? It's really not all that mysterious. Even though we use bipolar transistors for analog amplification, they make good DC switches too. A saturated bipolar device will saturate the 10-50ma of a pedal's drain to well under 100mV if you drive the base hard enough. A PNP connected in series with a battery can be turned on by pulling its base toward ground with a base-current limiting resistor, so it makes a good switch for turning + power on.

I guess I should have tried to patent that, given the dreck that's awarded patents, but it's really a standard trick with bipolar parts. Works with P-channel MOSFETs as well, and there you don't even need a limiting resistor. But you DO need a gate-source zener.