Lm317 current limit?

[Esppse]

Hello,

I'm trying to build a 4 tube version of this here

https://www.....org/download/file.php?id=5012&sid=bf03b8f32f2218001020653914daf50f&mode=view

I got 2 to work successfully, when I added the other 2, the voltage of the heaters went from 1.2v to 0.3. Is this because of a current limit of the LM317?

I don't see any short, the voltage to the plate of all the tubes is still the same, running at 33v

One difference is for the 1st 2 tubes I used 500k trimpots to bias, I used 200k for the last 2 because I ran out

[Rob Strand]

The heater spec is 1.2V @ 21mA to 26mA.       Unless I'm missing something.
So 4x heaters is 100mA. (forgot the 100 ohm but it doesn't change things)

Well below the current limit.

One possibility is thermal shutdown of the regulator:
Say the 9V rail is 12V, at 100mA the regulator will dissipate about 1.1W.
The regulator without a heatsink has Rth = 24 degC/W to 38degC/W
So we would expect to see less than 50 degC temperature rise.
So that's not the problem.

Another possibility is the 9V rail is dropping. If you have a 1A transformer then that's unlikely.
You should measure the "9V" unregulated DC input voltage to the regulator.

Looks like something is wrong.
Maybe measure the resistance on all your heaters.
Do it when they are all cold (all at the same temperature).
if you want to check the current they need to be hot (just measure the current of one when running).

[antonis]

Say the 9V rail is 12V, at 100mA the regulator will dissipate about 1.1W.

the voltage to the plate of all the tubes is still the same, running at 33v

Most possible is LM317 been fed directly from +33V..
If so, 2.4W dissipation could raise regulator temp 90^oC which added on enclosure working temperature (depending on tubes  heat emission) could reasult into thermal shutdown..

P.S.
@Esppse: Is LM317 heat tab connected on metal enclosure..??

edit: Just saw this: https://www.diystompboxes.com/smfforum/index.php?topic=127841.0
MAX1044 out current capability is 10mA..
Could you plz post a schematic fo your particular circuit..??

[Esppse]

Ok I have been struggling with this all night and found some things out.

When 2 heaters are on one regulator, the voltage is 1.2, normal.

As I add a third one, it drops down to 0.8.
With the fourth, it drops to 0 4.

So I set up a second LM317, which takes from the same 9V, and the heaters now are all at 1.25, 2 heaters per regulator.

Now the problem is that I am getting an terrible oscillation when putting the 3rd and 4th tubes in. Here's how I have it.

ZHBOOST (2 tubes) ---> AMT B1 ----> LAST 2 TUBES

The way I added a tube to the ZH boost is simply a 0.1 coupling cap between them. Same with the last 2 tubes. 0.1 cap, TUBE 3, 0.1 cap TUBE 4, 0.1 cap. The idea is to have a tubey response before and after solid state distortion. My goodness it sounds amazing, but the oscillation noise is an issue.

So I isolated these sections and they work on their own.
ZHBOOST ---AMT B1
LAST 2 TUBES

But when I put the last 2 tubes after everything, there are huge oscillation pulses.

I tried messing with input impedance before the last 2 tubes, as well as input cap values but it's still there.

Could the B+ voltage with is 33v be feeding back and going through the B1, creating a loop? Or is this the current limit, and I need a second charge pump for the last 2 tubes?

I feel I'm so close to solving this mystery

[antonis]

Could you plz post a schematic for your particular circuit..??

Let's take it once more..

[anotherjim]

If you measure one tube's filament resistance when it's totally at room temperature, divide it by 4 then you have the total load resistance the regulator faces when it powers on into four of them. It's seriously low. The difference between the hot & cold resistance of a filament can be immense!

You may notice that tube amps rarely have fused filament supplies. It's difficult to pick a rating for the fuse that protects against a fault but doesn't get blown by the cold resistance start current. I recently looked at a modern VOX AC30, and that had a blown fuse for the filaments. It was a 6.3A which should be ok for the working current but a blown fuse was all it was, I couldn't get it to fail again. The equivalent AC15 has fewer tubes, but the same 6.3A rating. The next choice above is 10A and if the AC30 blows the fuse again I'll stick a 10A in!

Anyway, you could try a delayed soft start in the regulator circuit.
This kind of thing...
http://www.circuitdiagram.org/lm317-soft-start-power-supply-adjustable.html

[Esppse]

Could you plz post a schematic for your particular circuit..??

Let's take it once more..

Ok here it is, I didn't have a schematic or the B1, though it works perfectly on its own. Below is the vero layout I used.

http://johnkvintageguitars.homestead.com/Effects/Fuzz-ODs/AMT/B1_Legend_Vero.png

[Rob Strand]

One thing still not clear is how you are generating the voltage on the input side of the LM317.

Do do have a transformer feeding a rectifier and filter cap like you the schematic you posted earlier on?
And is that transformer *only* powering the LM317 heater supply?
If heater supply is going elsewhere then you can get trouble.

[antonis]

One thing still not clear is how you are generating the voltage on the input side of the LM317.

I've a strong feeling about that..
https://www.diystompboxes.com/smfforum/index.php?topic=127841.0

[Esppse]

Actually it's just 9V to input pin of regulator, 100r between pin 2 and 3, pin 3 ground, and pin 2 is 1.25, just like the schematic from the original.

The 33v is coming from a charge pump, which draws from the same 9v source.

https://2.bp.blogspot.com/-DyDUlNDusXQ/WKO1Mv6HQuI/AAAAAAAAAzw/CmNP9ZcTr8YcMMEF25xUjyGVPW2DkbjdACLcB/s1600/33V%2BCharge%2BPump.png

[antonis]

Could you plz plz plz plz plz plz post a schematic of particular LM317 wiring..??
(what you said about regulator wiring makes no sense at all..)

[Rob Strand]

Could you plz plz plz plz plz plz post a schematic of particular LM317 wiring..??
(what you said about regulator wiring makes no sense at all..)

I didn't get it at first but what Esppse is saying is the 9V DC goes to the LM317 circuit and also to the 33V converter.   The whole circuit is wired with a common 0V.   The simplest simplest wiring and no rectifier or diodes etc.

I get anotherjims point about the start-up current.   Given the running heater current is in the order of 100mA for four tubes we probably wouldn't expect the start-up current to go over 1A (maybe 700mA or so).   With a 1A limit the heaters should warm enough to start raising the resistance and heater currents would be expected to stabilized normally.

It would be very worthwhile measuring the 9V current to the heater regulator for 1, 2, 3, 4, tubes (leave the caps connected the regulator input).   Also check the 9V rail isn't saging down with four tubes heaters.

Another point, if it works with tubes "1" and "2" connected does it work with tubes only "3" and "4" connected, just in case the circuit with tubes "3" or "4" has an issue (for whatever reason).

If it were mine I'd be removing all the tubes then placing resistor loads on the output of the 1.2V supply to see what current I can pull under "test" conditions.   You could place a load of 47 ohms, then two 47 ohms in parallel then three etc.  measuring the output voltage in each case.    If you get to 4x47ohms in parallel start using heavier loads to see what point the regulator voltage drops to 0.8V to 1V.

[antonis]

Actually it's just 9V to input pin of regulator, 100r between pin 2 and 3, pin 3 ground, and pin 2 is 1.25

Taking into account LM317 pinout (1=Adj, 2=Out, 3=In), it still makes no sense to me but let it be..