Forumvibe issue

[jkokura]

This is not my usual level of detail when it comes to debugging, but if needed I'll post all the pics and voltages and stuff. Brad's a good guy, and this was PM to him originally that he's asked me to post here:

I have a forumvibe I'm having trouble with, and your (Brad/RedHouse) post today reminded me that I wanted to ask for a little debugging help on it. I bought it from Barry on BYOC about a year ago now I think.

Anyway, It's from the forumvibe version one documentation. All the voltages look right according to that document. I've built it to forumvibe specs, not building the vintage vibe or 'redhouse mods' versions. You can see that document here: http://www.lynx.bc.ca/~jc/ForumVibe.pdf (Big PDF - be aware before clicking)

I've got signal, and I can vary the brightness of the bulb using the trimpots, but I'm getting no LFO action on the bulb. The bulb just says lit at whatever brightness the trimpots resistance sets. Before I do pictures or give any voltages or anything, is there anything you can recommend me take a look at? I've been over it many times and have, as far as I can see, included all parts in the correct places with the correct orientations and such. I've included all the jumpers, and I am using a copper pipe cap that hasn't been soldered in yet as a sheild for the LDR's. I bought the good LDR's and bulb from Smallbear.

Jacob

[RedHouse]

Ok thanks for the post Jacob, sorry it took me some time to get here, got hung up with some unpleasantness elsewhere.

My first though if your bulb is lighting but not pulsing, is to check these things:
(keep an open mind, no criticism intended here, just places to look)

• Is your speed pot hooked up? w/o a pot hooked up it will act just as you describe
  
• Is the little jumper (J5) installed? is it installed in the middle and bottom holes?
  (it's located between Q14 and TR3).
• Do you have the "Intensity" pot hooked up as shown on page-18 of the original PDF?
• Do you have the "Speed" pot hooked up as shown on page-18 of the original PDF?
• Double check Q11/Q12 and Q13/14 orientation, all emitters are to the left if looking at your board as the diagram on page 5 of the original PDF?
• Make sure the two 10uF caps (C22/C23) have their butts (-) attached to the Intensity pot leads

If the bulb lights it means the driver transistor(s) are working (Q13/14), it the light doesn't pulse/throb the place to look is where the LFO outputs the signal to the bulb driver transistors.

Do you happen to have an oscilloscope? if not you can use the Bob Sweet LED method for LFO detection, which is to temporarily lift the leg of R45 where it connects to the ground trace, solder in a couple pieces of wire so you can clip an LED onto them, clip an LED there with the cathode side to the ground trace (cathode is the shorter leg of an LED, or the lead with a flat spot on the LED's side) and the anode side clipped to the leg of R45 that is sticking up off the board now.

With this LED in place, the LED will glow with the tempo of the LFO regardless of the settings of the "Intensity" pot or bulb driver circuitry so you can isolate where the problem lies and get a visual indication (although poor) if the LFO is oscillating or not.

Try looking into these things and let me know what you find, if I don't see your post email me that you've posted a reply.

[jkokura]

My first though if your bulb is lighting but not pulsing, is to check these things:

• Is your speed pot hooked up? w/o a pot hooked up it will act just as you describe
  
• Is the little jumper (J5) installed? is it installed in the middle and bottom holes?
  (it's located between Q14 and TR3).
• Do you have the "Intensity" pot hooked up as shown on page-18 of the original PDF?
• Do you have the "Speed" pot hooked up as shown on page-18 of the original PDF?
• Double check Q11/Q12 and Q13/14 orientation, all emitters are to the left if looking at your board as the diagram on page 5 of the original PDF?
• Make sure the two 10uF caps (C22/C23) have their butts (-) attached to the Intensity pot leads

If the bulb lights it means the driver transistor(s) are working (Q13/14), it the light doesn't pulse/throb the place to look is where the LFO outputs the signal to the bulb driver transistors.

Do you happen to have an oscilloscope? if not you can use the Bob Sweet LED method for LFO detection, which is to temporarily lift the leg of R45 where it connects to the ground trace, solder in a couple pieces of wire so you can clip an LED onto them, clip an LED there with the cathode side to the ground trace (cathode is the shorter leg of an LED, or the lead with a flat spot on the LED's side) and the anode side clipped to the leg of R45 that is sticking up off the board now.

With this LED in place, the LED will glow with the tempo of the LFO regardless of the settings of the "Intensity" pot or bulb driver circuitry so you can isolate where the problem lies and get a visual indication (although poor) if the LFO is oscillating or not.

Try looking into these things and let me know what you find, if I don't see your post email me that you've posted a reply.

Thanks Brad. Here's my response:

1. Yep, speed pots hooked up and as far as I can tell it's right.
2. Yes, and I did put it in the lower two holes as you indicated
3. Yes, the intensity pot is hooked up correctly.
4. Yes the Speed pot is hooked up correctly
5. I have double checked the orientation, and they are oriented correctly. I'm using 2N3904's there, so perhaps the issue is that they aren't 5088/9's?
6. Yes, those caps are oriented correctly.

I'm perplexed to why the bulb is working, but not pulsing. Now that I've realized that I'm using 3904's maybe that's the culprit?

I don't have an oscilloscope, but I'm getting the feeling lately it might be worth looking into.

I'll try that method of lifting the leg or R45. do I need another LED in series, or will that one LED do? What if it doesn't light? What if it lights but doesn't pulse?

Let me know about the trannsistor sub, I would have thought the 3904 wouldn't be the issue...

Jacob

[R.G.]

Look at all the "Neovibe LFO doesn't oscillate" posts here.

The LFO in the Univibe, Neovibe, and to the extent they're identical, the forum vibe is heir to these problems:
1. Hookup and wiring. Gotta get the parts in the right place, right value, and correct orientation, including the pinout of the transistors.
2. Enough current gain for the LFO transistors. If the gain is too low, the typical response is that it blinks a couple of times when the power is turned on, then quits. The solution is more gain. High gain for the first transistor (I don't know how the part numbering varies to the forum vibe). A low-Vgsoff JFET like the J201 will work for the first transistor if you get the pinout right. So will subbing in a darlington like an MPSA13 or 14.
3. Pots. You can eliminate pot issues till you get it working by using two resistors instead of the pots. When it works, then wire up pots.
4. Caps. Polarity really matters for the three stacked 1uFs.

[jkokura]

Look at all the "Neovibe LFO doesn't oscillate" posts here.

The LFO in the Univibe, Neovibe, and to the extent they're identical, the forum vibe is heir to these problems:
1. Hookup and wiring. Gotta get the parts in the right place, right value, and correct orientation, including the pinout of the transistors.
2. Enough current gain for the LFO transistors. If the gain is too low, the typical response is that it blinks a couple of times when the power is turned on, then quits. The solution is more gain. High gain for the first transistor (I don't know how the part numbering varies to the forum vibe). A low-Vgsoff JFET like the J201 will work for the first transistor if you get the pinout right. So will subbing in a darlington like an MPSA13 or 14.
3. Pots. You can eliminate pot issues till you get it working by using two resistors instead of the pots. When it works, then wire up pots.
4. Caps. Polarity really matters for the three stacked 1uFs.

Thanks RG. Good to know that I can look up those other projects, thanks. Your wise advice and encouragement are something I cherish, so I hope you help me out some more.

1. I know that I'm fully susceptible to the common part and hookup wiring errors, and I'm not eliminating them entirely, but I'm no rookie to this dance. I've been trying to get this working for about 6 months, so I'm very certain I've eliminated all the usual part error bad wiring issues. I'm getting sound, and it's usually very good, I just can't get the bulb to blink, and my eyes and calculators haven't found any part errors as of yet. Today was the first time that I realized that perhaps the 3904's might be the issue, haven't gotten in there to switch them out for 5088's yet. The specs for the Forumvibe indicate that 2n3904's are totally fine in that position though, and the pinouts have been checked, but I might have a funky 2n3904 with a backwards pinout...

2. what would you guess the problem is if the bulb goes on and stays on instead of oscillating? Is that a gain issue as well, as you suggested having not enough gain would cause the bulb not to oscillate? I don't have any MPSA series trannies... Just the usual MOSFET, JFET and BiPolars.

3. I have verified the pots by replacing them with resistors and checking the pot's function. Not the issue as far as I can tell.

4. I have two polarized caps in there, C22 and C23 and they were rechecked by me and my wife today per Brad's suggestion.

I think I need to take some good pictures and post them. I can take measurements on the LFO trannies and post them as well. Perhaps I'll be able to do that tomorrow.

Jacob

[pinkjimiphoton]

after trying a mod i read about online with a dano chicken salad vibe, i can offer one small piece of advice that may be useful...cuz the mod i tried KILLED my vibe...
slowly adjust the bias or whatever that adjusts the intensity of the lamp...just a fraction of a turn can be enough where it just plain won't work, and after you adjust it, turn the effect off, and then back on again. for some reason, that seems to matter sometime. no idea why...just what i discovered messing around. maybe someone way more knowledgeable than me can elaborate on why that may work, but i've seen it happen on a few of my builds now. it's really weird. near as i can figure, there's a balance that needs to be met in the circuit, and tweaking the circuit can upset that balance. turning the effect off and on seems to re-set it or something.
voodoo? mojo? idiocy? beats me...but for some reason, it sure seems to help sometimes!!


btw...don't waste your time with that chicken salad mod that's out there. adding foil to the vactrol doesn't do poo, and adjusting the trim pot for the circuit is more trouble than it's worth, and can have you debugging for hours trying to get it right again.

[RedHouse]

My first though if your bulb is lighting but not pulsing, is to check these things:

• Is your speed pot hooked up? w/o a pot hooked up it will act just as you describe
  
• Is the little jumper (J5) installed? is it installed in the middle and bottom holes?
  (it's located between Q14 and TR3).
• Do you have the "Intensity" pot hooked up as shown on page-18 of the original PDF?
• Do you have the "Speed" pot hooked up as shown on page-18 of the original PDF?
• Double check Q11/Q12 and Q13/14 orientation, all emitters are to the left if looking at your board as the diagram on page 5 of the original PDF?
• Make sure the two 10uF caps (C22/C23) have their butts (-) attached to the Intensity pot leads

If the bulb lights it means the driver transistor(s) are working (Q13/14), it the light doesn't pulse/throb the place to look is where the LFO outputs the signal to the bulb driver transistors.

Do you happen to have an oscilloscope? if not you can use the Bob Sweet LED method for LFO detection, which is to temporarily lift the leg of R45 where it connects to the ground trace, solder in a couple pieces of wire so you can clip an LED onto them, clip an LED there with the cathode side to the ground trace (cathode is the shorter leg of an LED, or the lead with a flat spot on the LED's side) and the anode side clipped to the leg of R45 that is sticking up off the board now.

With this LED in place, the LED will glow with the tempo of the LFO regardless of the settings of the "Intensity" pot or bulb driver circuitry so you can isolate where the problem lies and get a visual indication (although poor) if the LFO is oscillating or not.

Try looking into these things and let me know what you find, if I don't see your post email me that you've posted a reply.

Thanks Brad. Here's my response:

1. Yep, speed pots hooked up and as far as I can tell it's right.
2. Yes, and I did put it in the lower two holes as you indicated
3. Yes, the intensity pot is hooked up correctly.
4. Yes the Speed pot is hooked up correctly
5. I have double checked the orientation, and they are oriented correctly. I'm using 2N3904's there, so perhaps the issue is that they aren't 5088/9's?
6. Yes, those caps are oriented correctly.

I'm perplexed to why the bulb is working, but not pulsing. Now that I've realized that I'm using 3904's maybe that's the culprit?

I don't have an oscilloscope, but I'm getting the feeling lately it might be worth looking into.

I'll try that method of lifting the leg or R45. do I need another LED in series, or will that one LED do? What if it doesn't light? What if it lights but doesn't pulse?

Let me know about the trannsistor sub, I would have thought the 3904 wouldn't be the issue...

Jacob

Ok, it's best practice to separate the systems and focus on the area that is failing.

The things that cause the bulb to operate correctly are basically 2 sections, an oscillator, and a bulb driver circuit. Everything before the 10uF cap that hooks to the "Intensity" pot's pin-3 (CW) is part of the oscillator, everything after the other 10uF cap that hooks to pin-2 (wiper) of the "Intensity" pot is part of the bulb driver circuit.

You have already said you can light the bulb, so I'd turn your attention to the oscillator section and see it if's oscillating, add that LED as I posted above and re-start the unit, you will know instantly if the oscillator is oscillating, the LED will glow at the speed your speed pot is set to, it is before the "Intensity" pot so it doesn't matter where you set that.

Rg made some good points above (than you for that RG) but please try the LED thing it's quick-n-easy and can tell you if your oscillator is working before you start replacing parts.

One other thing, if you used polar electrolytic capa, RG was right on the money with that tip about their orientation, I forgot about that as I usually use Poly or NP electro's anymore, haven't done a build with polar electro's for a long time, do check that.

Add the LED and tell me/us if it's pulsing.

[jkokura]

Alrighty.

So first off, I found I had added a 1uF cap above the 15nF cap. Removed that, no change. I think it's there in version 2, and not in version 1, so I put this back to version 1 specs.

Second off, I lifted the leg of R45 and hooked in an LED using a couple alligator clip wires. It's pulsing, and the speed knob is obviously working as well. So the LFO is working.

Third, I quadruple checked all my parts as well, each one at a time and checking them off on the layout. They are all perfectly in position.

Fourth I took some voltages. Q1-10 seem to be just about perfect as compared to the document's recorded values. No sense in posting them. Q11-14 showed some interesting results. They all seemed to be fairly normal, but especially on the two left most Q's, the voltage on the left pins were oscillating quite strongly. The other voltages seemed to be working fairly fine.

So I'm stumped. If I could audio probe, I'd know what the problem is, but since I'm getting fairly normal signal all the way through the circuit, I can't use that tool as normal! I don't have an osciliscope, nor can I afford to just go buy one, and I don't know if that would properly help me at all even. But the DMM has shown me things seem to be normal, so I don't know what to do now.

I should also note that these are the parts I used for the Bulb and LDR's.
Bulb: http://www.smallbearelec.com/Detail.bok?no=716
LDR: http://www.smallbearelec.com/Detail.bok?no=712

In the pictures below you can see that I am getting my +15V from a road rage board from Madbean. It supplies me the 15V by doubling the 9V using a charge pump and then using a 78L15 regulator to bring the voltage down. I have verified that it is supplying almost exactly 15V to the circuit, and that there is no power leaking to ground.



I don't mean to brag, but my solder job looks really good in this picture. My wife agrees.



Jacob

[RedHouse]

I agree with your wife, your soldering looks really good Jacob.

Quick-n-easy that whole LED thing isn't it, confirms oscillation in a few seconds.
(R.I.P.  Bob Sweet)

Ok so now that you confirmed your oscillator is indeed oscillating (good job BTW), try these things...

Turn your "Intensity" pot down all the way so the lead that is connected to C23 is grounded.

Turn TR3 and see if you can cause the bulb to go from OFF to ON, be careful not to go too far ON as you could burn-out your bulb. Under normal circumstances (when bulb bias is set at 1/3 V+) then TR3 can turn the bulb all the way off and all the way on, even on too far for bulbs lower than 12v.

(keep in mind you have your Intensity pot turned down, you have to turn it back up when you test a change and back down as you make new changes)

This next bit is harder but look closely at TR2, the pin on that trimmer that connects to J4 should measure 1/3 of it's total resistance, the wiper pin and the pin connected to B+ (and Q13's collector) should measure the remaining 66% of the total resistance of that trimmer (TR2), you need to get that trimmer set like that before moving on.

If you can't get it right just replace TR2 (temporarily) with the two resistors (R48, R49) as shown on the "vintage build" page of the PDF.
(the v1 boards this trimmer is physically installed backwards from where it should be (my bad, sorry) and so the 33% mark is 66% which is backwards from the way it was designed to be used with the Piher pots)

The whole point it to get the bulb biased to 1/3 V+ for further debugging. TR2 and TR3 are very interactive, you have to set them to the default place before adjusting them. If you never get them to their default place you will not get proper functionality from the LFO/Bulb-driver section of the vibe.

TR2 must be set right before you can set TR3 right, both need to be right before you can tweak things into their best settings.

Once you know you have the correct bulb bias, turn TR3 so it's glowing orange, at this point you should be able to turn up your Intensity pot and you should see the bulb start pulsing as normal operation.

[R.G.]

If you know the LFO is oscillating, how do you know:
1) the signal is getting to the bulb driver transistor
and
2) the bulb driver transistor is working? Yes, bulb intensity varies with the bulb bias pot, but it would do that if the transistor were hooked up wrong or defective.

You measure transistor voltages. Can you post the lamp driver voltages and the voltages on the depth pot?

[RedHouse]

Yeah, but the voltages won't be meaningful until he gets that bulb bias set back to default (100k/47k), I'm not disagreeing with RG on this but my gut feeling is that TR2 is where the problem is (has been before for others) but certainly a defective or otherwise misbehaving trasnsitor can do it too so carry-on and lets see where we go.

The bias offset mod trimmer can swamp the driver transistor into saturation which will light the bulb steady Jacob describes it doing, I'd encourage him to get the thing into the proper parameters first.

It's this particular implementation that lacks (single trimpot across the base) it would have been better to just keep the 47k resistor and add a 100k trimmer and another resistor (5-10K?) on the top end of the trimmer to confine the range from default bias to just this side of saturation.   

[R.G.]

Yeah, but the voltages won't be meaningful until he gets that bulb bias set back to default

Maybe. It may be just me, but knowing where the DC is - especially when the problem is that the AC is not showing up where it should be - has always helped me with debugging before. Call me sentimental...  

Edit: I guess I really should go look at the schemo, layout, etc. I have never studied the forum vibe schemo or layout in detail. I guess I should.

[RedHouse]

...I guess I really should go look at the schemo, layout, etc. I have never studied the forum vibe schemo or layout in detail. I guess I should.

The doc is offline at the moment (recall that "other" thread recently), if you feel like sending me a PM me with an email address, I'd be happy to sent you the layout and schem pages of the newest v3 doc I'm putting together.

[RedHouse]

Any progress Jacob?

[jkokura]

Brad! Yikes it's been a long time since I worked on this.

I had a big contract this summer and put this down back in May. Getting back I haven't picked it up yet.

I have a theory that the problem is that the TC1044 charge pump isn't supplying enough current to the circuit - I think the specs are just 20ma come off the TC1044, so even though it's the right voltage the current isn't enough to drive the transistors for the bulb. I have an LT1054 now, and I will be plopping that into a charge pump board - it's specs say I should be able to draw 100ma off that pump, which I imagine will solve the problem.

Jacob

[RedHouse]

Brad! Yikes it's been a long time since I worked on this.

I had a big contract this summer and put this down back in May. Getting back I haven't picked it up yet.

I have a theory that the problem is that the TC1044 charge pump isn't supplying enough current to the circuit - I think the specs are just 20ma come off the TC1044, so even though it's the right voltage the current isn't enough to drive the transistors for the bulb. I have an LT1054 now, and I will be plopping that into a charge pump board - it's specs say I should be able to draw 100ma off that pump, which I imagine will solve the problem.

Jacob

Yeah, 20mA is likely a problem, the bulb itself can swallow that right off the top.

[Jaicen_solo]

Agreed, that would be the first thing to try.
That said, are you powering the LFO with 15v? On R.G's board, the LFO runs off the DC input, which is what, 22v?
Does the Forumvibe have the 15v regulator onboard? Don't know if it would make much difference, but why not hook up a PSU or two 9v batteries in series and see if that works?

[Barcode80]

...I guess I really should go look at the schemo, layout, etc. I have never studied the forum vibe schemo or layout in detail. I guess I should.

The doc is offline at the moment (recall that "other" thread recently), if you feel like sending me a PM me with an email address, I'd be happy to sent you the layout and schem pages of the newest v3 doc I'm putting together.

What "other" thread?

[RedHouse]

What "other" thread?

You realize the post you're quoting was more than 6 months old right? anyway, it's all water under the bridge, everyone has moved on as far as I  know.

Doc's can be found here: www.classicamplification.net/forumvibe

[R.G.]

People tend to us a lot of bulbs of different provenances in that circuit.

A bulb is a variable resistor. It heats up, and as it heats, its resistance goes up. Hot resistance may be 10X the cold resistance for some bulbs. The univibe driver tends to put a quasi-current-source on the bulb. It's not pure, because there is some feedback from the collector trying to bring the output impedance down. But current through the bulb doesn't vary anything like linearly with the available power supply voltage there; as witness that people use an incredible number of different power supplies on their incredible number of different bulbs.

Going to the LT part would be a good idea. Give the bulb all the current it wants. It'll be happier. Watch for secondary issues like the charge pump converter sucking pulses of current on the ground wire and causing whine issues. Putting a BFC across the charge pump right at the pump so it can suck current locally instead of wiggling the power supply lines is also a good idea, as is putting some 0.1 and 0.01uF ceramics close to the charge pump to keep the impedance into the pump low at RF as well.