Menu

Show posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Show posts Menu

Messages - Roly

#76
"Runcible" is a nonsense word invented by poet Edward Lear and first used in "The Owl And The Pussy Cat", but Lear never defined what it meant and used it in other places where the meanings were inconsistent, just for the sound.

In my use it means an arrangement that could be better and more refined, not impressive, fairly ordinary or lacking.
#77
Quote from: JLTcouldn't handle the power

Which is likely to be the case with just about any consumer/domestic speaker these days that you pump a genuine 15Wrms into.

The preamp has a problem, very low input impedance, only about 60k without the optional 10k killer resistor.  Passive guitars don't like.  You need at least a buffer of some sort (e.g. FET) to get it up to at least 1Meg.

It also could have more gain than you need and you can moderate it by increasing the value of R6/100 ohms to say 220 ohms.

HTH
#78


Found a cleaner print;




Sadly the thread on tdpri.com doesn't have much to offer, only a long rambling debate about transistors, distortion, and negative feedback.



SWTPC - Southwest Technical Products Corp. (a.k.a. "sweat pack") had a good reputation in the day, did a lot of very early personal microcomputer stuff, S-100 buss.




The "22 watts" is IHFM or "music power", the actual rating of the amp is 18 watts RMS (which is typical of modern small gigging combos and generally more than enough).

It claims a very high bandwidth, 20Hz to 100kHz, which is fine but would need to be confined for guitar work. (or excessive hiss)

The raw input sensitivity is 1.5V for full output.  A guitar will make a noise but not drive it fully.

The raw input impedance is more of a problem at only 5k.  This is fine for a power amp but far too low for an unbuffered (passive) guitar.

So you are going to need a preamp of some sort to provide;

- an input impedance of 1Meg or more
- voltage gain for sensitivity
- EQ/tonestack for sound colouration
- a volume control

The very simplest thing you can do is put a FET source-follower buffer in front of it, but it really needs a preamp that is a bit more full-blooded.  But later.


On closer consideration...

Uh oh ... "one silicon and one germanium output transistor" (and no emitter resistors either)

That does not bode well.

Sorry, but I don't think this circuit makes a good starting point.




PS.

Quote from: cbg Rick1. How does one decide what Ic (collector current) value to use for each stage of an amp? The information I'm using says to use an Ic of 5ma for the first transistor and an Ic of 10 ma for the second transistor. Just wondering how these values were arrived at  ???

I go about it backwards.  For 50W in 8 ohms we need about 5 amps peak.  Assuming a venerable 2N3055 output transistor, it has a current gain, Hfe, of as low as 20 at this current.  This means that it will require;

5/20 = 0.25A into the base.

Now the driver has to deliver 250mA peak, and assuming an equally venerable BD139 as the driver it has a minimum Hfe of 40, so it will in turn require;

250/40 = 6.25mA of drive at its base.

And so it goes.

For preamp and small signal stages one can assume a current of about 1mA which is about where most small signal transistor parameters give their best, gain, noise, but this is a massive generalisation and the datasheet always rules.


Quote from: cbg Rick2. In the schematic above I was able to come up with values for the collector and emitter resistors for the first transistor and for the resistor from base to ground..... but then I noticed there is no direct DC current path from the base to the positive rail. How do I calculate the value of that resistance?

The base current for the first transistor is derived from the output "half-rail" (emitters of Q4 and Q5).

At power-on Q3 and Q5 start to conduct until the half-rail rises to where Q1 brings the rise under control.

This bias is also therefore a DC (and AC) negative feedback path from the output.  The pot R5 is used to adjust the half-rail so it idles at half the voltage of the supply (and the AC gain ends up being whatever it is at that setting - a runcible arrangement IMO).
#79

R87  0V
R88  0V
R89  1.5mV
R90  14mV

R102 45mV
R104 47mV
R106 1.03V

1.5mV across R89, but 0V across R87.

If the current through R89 isn't coming through R87 where is it coming from?   ???

On the strength of these voltages I'd test Q11 out of circuit ... but I'm dubious because it seems like a rather odd failure mode - an open emitter with a leaky collector.  Not impossible, but ...


{thinking out loud;
In this amp the natural state of the upper and lower output "cells" is on (possibly the reason for the excessive protection circuitry, Q12-17).

The upper cell is turned on by R91 and R91 and this current injection is shunted through D28.

Simlarly the lower cell is turned on by R99 and R100 and this current injection is shunted through D29.

These voltages suggest that there is nothing wrong with the output transistors or drivers, but things get strange when we get back to the pre-driver Q11.}


I must confess that I'm at a bit of a loss at the moment.

BRAINS TRUST!  HALP!  Gitin'ere, we need more smarts on this one.


brodie1600 please double check your voltage measurement across R87.

Also please measure and post the voltage to ground at the collectors of Q9, 10, and 11.

#80
Good ol' Hunner - tripewriter, shotgun, joint...


HST's remains being blasted out of a cannon.

Quote"I hate to advocate drugs, alcohol, violence or insanity to anyone, but they've always worked for me" - HST

They do not, however, work so well when fixing amps.   8|
#81
The Newcomer's Forum / Re: Old newbee
April 30, 2015, 06:57:05 AM
Hi cbg Rick, welcome back to the fold.


Oh I do like these scratch instruments and amps, creativity in action.

As Enzo says, post the circuit, or link to, and we'll drag it right up to the current minute.   :dbtu:
#82
Quote from: dogfaceHey really enjoying this forum, already got some great advice and can't believe what a great place this is.

And worth every penny.   8|

Quote from: dogfaceBXR 400 I used as a gigging amp for several years. I've always wondered if this two power amp Bi-amp for two different speakers setup could be readily bridged

Don't know about the BXR 400 specifically (and I can't find a circuit) but generally bi-amps differ from stereo rack amps in that the two power amplifiers are not identical, each crafted to its intended bandwidth and speaker load, so they would not normally be bridge-able - they need to be identical, and even then it can be problematic.

A circuit will tell all.
#83
Hi dogface, welcome.


Quote from: dogfaceI found a thin brass mini switch nut of bigger diameter than the screw head, put everything back together, fired that beggar up and the ground hum was gone. To say I was tickled doesn't quite get it.

Good observation, well spotted, well fixed.  Enjoy to the full (it doesn't always go that easily)   :dbtu:

For all the blather on the forums advising everybody to "change all the caps", whatever the problem, ...

"I broke a string!"
"Change all the caps".

... this is the kind of stupid fault that is the bread and butter for amp techs.  It's not all loud bangs and clouds of smoke ya know.   ;)

Quote from: dogfaceI noticed that there was some residue around the cap leads ... does this shed any light ?

No.  Solder flux.  Quite normal.  Some companies wash it off, others don't.

Quote from: dogfaceI have a Fender BXR 400, a Kasino Concert, and a Laney Linebacker 100 watt chorus combo waiting in the wings all with issues I'd like to solve. 8)

Roll 'em out, one per thread (and I'd suggest one at a time to avoid confusion).
#84
Quote from: medium85So in your opinion this mod is a waste of time and money, or is it just the quality of the component that doesnt matter?

By far the most important thing is the value of the cap.  If you change a cap value, particularly in the tonestack, then the response is going to change in some way in proportion and you are likely to hear it.  But if you simply change all the caps for different dielectrics and the same values you are not going to hear it (whatever they claim - dielectric effects are real but truly microscopic in the context of a guitar amp).

{a great little free fun tool for playing with tonestacks is Duncan's Tone Stack Calculator.}

I am both a tech and a musician and I use both test instruments and my ears.  I can see distortion on my oscilloscope and I can measure it with my distortion meter, but I need my ears to tell me if it's nice or nasty sounding.

I know that my trained ear can just detect about 1% Total Harmonic Distortion (and about half that of crossover distortion), but my meter can measure down to at least 1/100th of that, way below what I can hear.

"Audiophiles" worry about distortion caused by capacitor dielectrics and try to eliminate them entirely, and claim they can clearly hear these effects.  The problem is that tests on various modern capacitor dielectrics with top end equipment cannot find any signs of distortion products down to 0.001%, so "hearing" something smaller than this has to be an illusion, a self-deception.

{in a dead silent environment the lower limit of hearing is due to the noise of blood pulsing near your ears, your heartbeat, and breathing.  And that's only if you have never played loud and got some Tinnitus, then that becomes your noise floor.  :(


}


The quality of components carried by reputable companies is more than good enough for any use in a guitar amp (or any amp for that matter).  The suggestion that there is something "wrong" with inexpensive over the counter parts, and that you need to spend a lot of money on special "mojo" capacitors with fancy names, is simply bunkum.  There is nothing wrong with Orange Drop caps and the like, it's just that they are normally vastly over priced for what they are.  {an exception are paper-in-oil which I wouldn't fit to anything.}


Quote from: medium85I have never listned to a modded amp, i have only heard youtube clips, and the sound is different than what my amp is putting out.
Could that be because they changed the speaker (and tubes) and that it is recorded through a mic?

Changing the cap values will have the biggest effect, changing the speaker may have some effect, changing the valves/tubes will have little or no effect.

Then it is where the clip was recorded and with what microphone/recorder.


If you want to experiment with changing some cap values, then just buy the values you need that your local electronics store stocks, change them, and use you own ears in your own space to evaluate the changes.  Don't like it?  Then just put back the components you changed.
#85
Hi medium85, welcome.


1. please post a link to the mod page so we can see what you are talking about and get a better idea of what you are trying to do.

2. silvered-mica capacitors are generally high quality and used in radio frequency circuits because of their stability.  Generally this level of stability is not required in audio circuits and polyprop or in fact just about any other dielectric, with the exception of ceramic, should do just fine.  Ceramic can be used and often are as a cheap substitute but they can also be microphonic.


A lot of what is written with great authority about capacitors and the effects that their dielectrics have on "tone" is imaginary rubbish intended to sell cheap components expensively.  We call these people cork-sniffers.




#86
Wot dem gentlemines said.   :dbtu:

I'm not real thrilled by that soldering either.  Dirty wave solder bath I'd guess.



The two joints arrowed are typical of the sort of cracks that can occur with larger components (or external sockets) where the joins can get strained and then crack;


The joints on the main filter caps don't appear to be cracked, but they don't look too darn hot either.

The scarring around C75/76 look like a bit of rough rework to me, perhaps somebody already replaced these at some point (but I suspect they were on the wrong tram), or maybe just a rough operator post-solder wave fit.

No, I don't think that's a short either, but check as Enzo advises.


So what's going on with your voltage readings?

The DC readings seem okay but the AC readings on the same points are way off.

The reason is that most less expensive meters cannot read AC accurately when there is also DC present, they see the DC as a much higher AC voltage.

The way to measure AC ripple on a DC rail with such a meter is to use a small cap, say 0.1uF, in series with the meter probe.  This blocks the DC but passes just the AC to the meter.


However, given;
Quote from: brodie160042 VDC / 92 VAC across both C71 and C72
... I don't think there is a crack on either cap, this is what I'd expect if the PSU was okay, and that the problem lies elsewhere.




A snippit of theory;
Quote from: brodie1600I was also unaware that there should only be DC flowing through the caps.

Ah, no, caps are insulators so DC doesn't flow through them.

These main filter caps also used to be called "reservoir" caps because they act as storage pools for electrons.  Every half-cycle of the mains the rectifier dumps a bucketload of electrons into the cap and its voltage rises a bit (Q = CV).  The cap provides a steady outflow of electrons to the output stage (Q = It).  Therefore there will always be some voltage ripple during this charge/discharge process, depending on the size of the "reservoir", bigger cap = less ripple (but it never gets to zero ripple no matter how big the caps are).


(this is either half of your power supply.  The dotted line shows the output if the cap is disconnected, pretty serious hum)

These charge/discharge cycles do produce an AC current through the cap we call ripple current which works out to be the same as the average DC output current.  If you want to investigate the physics of how an AC current can flow through an insulator look for capacitor displacement current.




Very loud hum doesn't automatically mean that there is something wrong with the power supply.  This can also be caused by the output stage drawing excessive current due to a fault such as a blown transistor.  This is why we all said "no" when you asked;
Quote from: brodie1600would it be safe to say that replacing all four caps would solve the issue?

Unless it turns out to be something simple such as a dud solder joint (which are fairly common) the output stage has many more components and has to be a serious suspect.


{Meanwhile, in a parallel Universe; if you had followed the advice of "someone" you would now be down two large caps, two small caps, four heavy diodes, and the time to fit them, none of which you can ethically charge to the client because there was nothing wrong with them - dead loss.  And you've still got the fault!}


(Movin' right along...) So we've got -1.19V on the output now as our main (and only) symptom.

{did I say I wasn't mad about this output stage?  :( }

This circuit contains a number of fusible resistors marked "Fu<value>".  In the pre-driver R87, R88, R89, R90.  in the driver R102, R104, R106.  These would be the next obvious and easy thing to check.

You can try measuring the resistance of each in situ and should get reasonable results, but it would be more useful to us if you could make up a list of the voltages at either end of each of these resistors.  That will give us a scatter of voltages across the output stage which may help to paint the picture.

Test conditions; all controls minimum, speaker disconnected.
#87
Quote from: mladenuhow can i mod my power amp to have straight current feedback?

{This is JMF's dance, but I'll just remark that} I don't see how you can remove the voltage feedback.

These s.s. power output stages are like power op-amps, they have very high open-loop forward voltage gain, let's guess at 50dB.  They apply about 30dB of Negative FeedBack to reduce the working voltage gain to about 20dB.

Just as the open loop gain of an op-amp without NFB would make it unworkable as a linear amplifier, so if you remove the voltage NFB from a power stage like this the input sensitivity will rise to the open loop figure, i.e. by 30dB to 50dB, or around 30 times greater.

As it is;

Quote from: J M FaheyFor full power it needs around 1V RMS , which is the unwritten standard.

So the sensitivity without VNFB would be;

1/30 = 0.03333333 or 30mV for full output.

But it's not just input sensitivity, the VNFB also reduces distortion by 30dB, significantly flattens the bandwidth, and prevents instability.  Without VNFB it is quite likely that the amp will idle with the output at one of the supply rails (across the speaker!) because there is now no mechanism for it to rebalanced the amp output to the zero volt reference at the input (the diff pair stage TR1/2).  Thermal drift will no longer be compensated.

The idea behind current feedback in s.s. amps is this;

- valve amps have a high source impedance to the speaker, in the order of ohms, so the dynamic damping of the speaker cone is low and the speaker is free to colour the sound.

- s.s. amps have a very low source impedance, small fractions of an ohm (in part due to the very high levels of NFB used) and so provide maximal dynamic damping to the speaker cone.

The effect of applying current feedback is that it causes the output impedance to be higher and give less dynamic damping, hopefully giving more of a "valve sound".

By the nature of s.s. output stages this must be an add-on to voltage feedback, and cannot substitute for it.
#88
Quote from: Enzobut you found DC on the output

I'll wait for that, and the supply voltages, to be confirmed by actual measurement with a voltmeter - it's only implied at this stage.
#89
Awwww... that is so cute, an LM386 with current feedback!   <3)



Lookit the lil' fella, pretending to be a Marshall stack.   :lmao:


{when these were new and came in round cans I tried quite a few things including power oscillator and compressor/limiter; useful but the signal quality isn't great.}
#90
Quote from: LoudthudCheck for a broken solder joint on one of the filter capacitors.

Quote from: EnzoI agree with LT, a very likely problem is cracked solder on one or both of the main filter caps.

Agreed - a simple test for a common problem.

- bright lamp and a lens
- identify the solder connections for C71 and C72 (look okay, or is one cracked?)
- gently push on each cap as you look at the joints with the lens
- any movement means it's bad and needs to be re-soldered.

If so that's likely to be the end of the story.

If not;
Get your meter out and measure the DC voltage across the output and post it here.


Quote from: brodie1600Someone told me it could be caps 71/72 or 75/76 that need to be replaced, or D53-56 might be shorted. Since I know that the DC is the issue, is there any other component in the schematic besides those that might be the culprit?

"Someone"?  Caps C71 and C72 are where we are telling you to check for cracked solder joints first.  That is more likely than one of them being duff ('tho that's still possible, but we check the simple main chances first).

C75 and C76 are the bypass caps for the preamp, but your someone has overlooked, or doesn't understand, the role of these caps or zeners D56 and D58.  The most likely result of either of these caps going open wouldn't be hum, it would be preamp instability, oscillations, putt-putt motorboating.

D53-D56 are the main bridge rectifier diodes.  Typically it's one goes and they all go, puts a dead short circuit across the power transformer which then either blows fuses or suffers a rapid smoking death.  If a rectifier is shorted it doesn't rectify any more so you don't get DC.  Again someone who has never actually repaired an amp with a shorted rectifier and is having wild shots in the dark.

Other components?  Sure, where would you like to start?  The output transistors, the driver transistors, the pre-driver transistors ... but they could all be innocent and it's just a cracked joint on C71 or C72.  This is a particularly complicated power amp circuit with quite a lot of protection circuitry, excessive in my view, and will take some care and accuracy to service, but it's possible because I did a curly fault in one of these by e-mail earlier in the year.

We will get to the transistors if that's where the symptoms lead, but first we verify the power supply, the first step in all repairs - "have we got good power?".

Quote from: brodie1600would it be safe to say that replacing all four caps would solve the issue?

Quote from: DrGonz78It is not safe to say that replacing the filter caps will fix anything at this point.

No it would not be safe to say.

This is called "scattergun" servicing, picking components at random and replacing them, a favorite of the clueless forums but it doesn't work.  Professional techs (who you are talking to here) first diagnose the fault, then replace the faulty components, promptly and accurately repair the amp, and stay in business.


Investigate the main cap solder joints and report findings.   :dbtu: