Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 09:28:47 AM

Login with username, password and session length

Recent Posts

 

Newcomer, Melbourne, Australia

Started by nodz, March 22, 2013, 04:42:53 PM

Previous topic - Next topic

nodz

#45
Hi Roly,

Also, if you have time, would it be possible for you to do an LTSpice on the PSU, I have modified the PSU to have only two points, for the two stages of the preamp
I have done one but I am not sure of the resistors I have in place.  What I'm actually struggling with is how to load each stage in LTSpice, so that I get an indicative voltage.  I've put a 1K resistor to GND at each filter stage.  Is this correct?

Cheers
Chris

Roly

The voltages in LTSpice are a bit of a trap - they are the peak voltages, not the RMS values.

Well if you want to model it under around 300mA load current 1k would be fine, but that's an awful lot for a small amp and I would be inclined to apply Ohms Law to the maximum currents you expect each section to draw and select load resistors of those values.

Your series resistors R2 and R5 look a bit on the low side.  In a low power amp these would typically be in the range 1-10k.  This should also reduce the hum levels on the preamp HT lines.
If you say theory and practice don't agree you haven't applied enough theory.

nodz

Ah right that explains it.  From the circuit diagram of my amp.  Wishing to feed the 150k plate resistors with 250v rms or approx 350v peak.  The amp was designed to drop about 100v across the plate resistor so the tube was fed at about 150v.  Therefore plate current is about 100/150k or 0.7mA.  This checks out on the valve curves using grid voltage of -1.25v and a cathode resistor of 1k8, gives voltage of 145v and a current of 0.7mA.

Now on to the spice load resistors.  Require 250v at 0.7mA which means they should be 357k.  Selecting nearest standard value of 330k to load each stage, I can change the values of the filter resistors R2 and R5 to get 350v peak on the traces.

Roly

You only need to use standard values for the real series resistors.  The resistors that simulate the draw of the preamp stages can and should be set for the current at voltage, i.e. 357k.

The most direct way to find the required resistors however is to add the DC portions of the triode stages to your power supply sim.
If you say theory and practice don't agree you haven't applied enough theory.

nodz

RIghtho, Just wired up the PSU for the preamp as shown in the diagram.  All voltages as measured as shown also.
All seems to check out except for the heaters.  I was expecting a calculated 6.34V from the LM317T as per spec sheet, that is,  1.25 x (R2/R1 + 1).
R2 is 1100 ohms (a 1K and 100R in series), R1 = 270 ohms.
No matter what I do, all I get is 1.247V, which to me, doesn't seem correct, effectively R2/R1 appears to be 0???

Pin 1 is the adjust
Pin 2 is the output
Pin 3 is the input

My understanding this was from left to right when the device was viewed from the front, i.e the plastic part closest to you and the metal tab at the back furthest away from you.

Can anyone offer any help please??

Cheers
Chris

Roly

That's looking pretty good.  :dbtu:

1.25 x (R2/R1 + 1)

1.25 * (1100/270 + 1) = 6.3425925  - Check!

I've never actually used one of these, but the same idea applied to a fixed regulator such as a 5 volt is that a current I flows in your R1 (270r) given by the 5 volts from output to common, and this same current flows in the grounded resistor, R2, giving an elevation of I * R2 volts (to which the regulator output is added giving the actual output voltage).

In the case of this chip the nominal output appears to be 1.25 volts, so;

I = 1.25/270 = 0.00462963 or about 4.6mA

elevation = 0.0046 * 1100 = 5.06 +1.25 = 6.31 - confirmed.

Your understanding of the pinout seems correct.

With only 1.25V out it is behaving as if there is no current flowing in R1 and R2, that the Adj pin is simply grounded (via R2).

Since the circuit and your calculations seem to be correct there is really no other option than to check the actual build very carefully for silly stuff like a dry joint on R1, R1 cracked, misread a 270k as 270 ohms, &c.

BTW, since you are running the heaters on DC you could run them in series at 12.6 volts and reduce the stress on the regulator, half the current and less voltage across it (if it's getting a bit too hot).
If you say theory and practice don't agree you haven't applied enough theory.

nodz

Appears to be a problem with R2 supposed to be 1K and 100R in series.
When I measure R2a its 100R, when I measure R2b it's also 100R even though it's marked 1K.  When I measure the resistance across both R2a and R2b expecting 1K1 I get 0 ohms.  Something very fishy.

Roly

Yeah, you've got a ground fault on the Adjust pin.
If you say theory and practice don't agree you haven't applied enough theory.

nodz

#53
Rightho,
Will have to check over, especially around that area
PSU is on veroboard.
Attached diagram shows how it is wired
Anything obvious from that?
Cheers,
Chris

phatt


Unless I'm sadly mistaken 4,700uF is way overkill and if my memory serves me I did read that large cap values on the output can actually blow the reg. :o
The DC has just been filtered by the reg so a large cap is not needed.
I believe a 10uF will do the job.

Another thought is why 6 volt?
You have 22 volts to play with,, I'd run the filaments at 12 volts.
just my 2 cents.

Phil.

nodz

#55
Most of the ideas are taken from audiophile circuits.  Pretty much a standard circuit for hi-fi preamp.  Perhaps may be overkill for a guitar amp but was potentially going to use the PSU for a hi-fi preamp as well.  Might try losing the voltage elevation link for the moment and just working on the heater by itself. I can change out the 270 resistor and replace with a 220 and I can just have 2 x 1k resistors in series for R2, voila 12.61volts.   I''ll see what else I have in my junk box by way of a smaller cap with approx 25wv to replace the 4700uF.

Roly

Quote from: nodzAudiophile

Ah, well there's ya problem right there.

Quote from: nodzAudiophile reasoning goes something like large capacitors for increased filtering, 6.3 volts and elevated to reduce hum.

Electronic engineering reasoning goes something like this.

The object of a bypass or filter cap is to reduce the source impedance of the supply.  The output impedance of a three pin regulator is already a fraction of an ohm, so putting an ohm or two in parallel with that in the form of a big electrolytic isn't going to make much difference.

What it will do however is store charge when the supply is turned off, and lacking a protection diode across the regulator will potentially drive current backwards through it and damage it.

The idea that a 6.3V heater circuit will have less hum is based in the wrong idea that the hum cause by the heater circuit is due to voltage.  In fact it is due to the magnetic field arising from the current, so a higher voltage heater will have less hum field.

But it's all pretty moot when the heater circuit is DC and doesn't have any hum field to start with.  A similar consideration applies to elevating the heater circuit above earth - if the heater is DC then there can be no AC leakage currents to the cathode.

The sad fact is that Audiophilia is an affliction, a kind of cult where the blind lead the blind.  Most of the concepts I've encountered from audiophiles have a tiny grain of truth at the origin, such as skin effect in cables, capacitor dielectric relaxation, and a hundred others, but are then expanded and misapplied by people who don't/can't get their facts right and in proportion.

Much of this codswallop sounds plausible, like the little copper oxide diodes behind OFC, but like that most of it is high class tommyrot.  There are people who are trying to be helpful, but there are also lots of sharks in the water making a good living off bulldust like gold-plated mains cables ot $1000+/metre speaker cables which are basically a con.
If you say theory and practice don't agree you haven't applied enough theory.

nodz

#57
Rightho changing heaters to 12.6v. I did have a diode across the lm317t but I thought that this may have been the cause of the problem so I removed it.  Will change the capacitance to something like 10uF, will see what I have.  Will remove the elevation link then when put the circuit board into the case ensure that both circuits are earthed properly.

nodz

#58
Woot, woot.  Preamp PSU completed.  R2 - 2 x 1K2 in series, R1 - 270R for a calculated 12.36V.  Actual measured voltage 12.5V with no load.
Changed out regulator, just in case.
Replaced the second 4,700/50V cap with a 100uF/25V, no problems  :tu:
Placed a diode across I/P & O/P, may not be required now that large cap has been replaced but for the sake of a 2 cent diode.
Simplified circuit diagram attached.

Board is larger than probably necessary, but I deliberately spaced everything out for ease of trouble shooting and repair, should it ever be required.

Tested supply on heaters of an ECC83 valve.  Nice glow.  :cheesy:

Thanks for the input, definitely some head scratching there for a while.

Now to start rounding up all the parts for the preamp.

Roly

Chances are if you fully account for the operating current from the regulator itself your voltage will work out exactly, but 12.5V measured for a required 12.6V is such a small difference it doesn't matter.   :dbtu:

The protection diodes may never come into play, but as you say, since they can be recovered from old gear free, why not?

A lot of builders make the mistake of miniaturising their board build, then putting it into a large case  :loco - just making life difficult for themselves.
If you say theory and practice don't agree you haven't applied enough theory.