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

March 28, 2024, 09:11:55 AM

Login with username, password and session length

Recent Posts

 

Hot rodding a simple SSPoweramp

Started by phatt, June 11, 2015, 05:30:32 AM

Previous topic - Next topic

phatt

Modifying My old Laney Amp which I've owned now for ~10 Years.

It's been a great little work horse which has been flogged to death but yet never given much trouble.
The power amp never gets Hot no matter how hard it's driven, luke warm at best even on the hottest days of summer. 8) 8) 8)
But I feel it could deliver just a bit more volume at times when gigs get loud.
Anyway rather than take a chance on a quick and dirty hot rod trick I thought I'd build a test circuit first.

The test has shown R11 (10k) feedback resistor can be doubled without issue which reaps an obvious and useful jump in volume and does not upset the circuit, exactly what I had hoped for.  <3)

The Laney transformer delivers 35-0-35 VDC with 2x 3,300uF caps while the test circuit is driven from 32-0-32 VDC but has slightly larger transformer and filtered with 6,800Uf caps.

Except for the 2N3442 power transistors the test circuit components are much the same, I/m using MJ15015 which have more power. I'm assuming they won't alter the result much?
2N3442,  140V- 10A- 117W
MJ15015, 120V- 15A- 180W

The test circuit is maintaining stable measurements and supply rails only dropping 2 volts at full power.
With audio signal input after 2 hours at full volume into a soak the heat sink still refuses to get past luke warm.  The tabs on Q3,5, and 6 are also only just warm. R9 (2k2) is 1 Watt also warm which I expected. Yay 2 wins in one day <3)

Of course a scope would come in handy right now but I've never owned one and hardly worth it for the little use it would get from me.
I just simulate this stuff and take some intelligent readings on the test bed.
I don't have all the correct models for my simulation program so it maybe a bit skewed but
a simulation shows doubling the FB resistor from 10k to 22k the output wattage jumps from 5 Watts to 20 Watts which seems a lot but the actual test Amp seems to cope fine with clear signal up until 3/4 volume when distortion starts to become obvious.   

All going well so far but a couple of things bother me which I'd like to resolve.

1/ While cross referencing actual Amp PCB to Schematic it seems there might be an issue with D7 and D9,, something not right? (marked on Drawing attach)
D9 looks wrong on the drawing yet is reversed on the actual board, which I think is Right.
While D7 seems drawn right on schematic but is reversed on board, which I think is Wrong. Head scratch? :duh

(This is the over current section and although it does not function in normal use I'd like to make sure it's correct in case of a short circuit.)

2/ The bias trim pot;
Am I right in assuming that, IF the wiper ever goes open circuit then the amp will likely self destruct.
If so, is there a better way to wire up the bias pot so that the Amp runs cold bias if the the pot fails.

Any help with that will be great.

My Amp model is a KD50 but the only schematic I found was for the KB80 but they are very similar.
Note 1; I have no idea why R2 (100k) is there but it just goes to ground and seems to do nothing.
Note 2; My test circuit does not have the protection system installed as it is not needed for this test.
Phil.

Enzo

You have the bias pot backwards.  If the wiper opens, it goes to full resistance, which turns on Q4 hard, making the amp coldest possible.  In fact if the pot itself opens, R14 pulls Q4 on fully with same result.  Coldest bias you can get is shorting across Q4, something we occasionally do during troubleshooting.  The only "harm" to the amp under those conditions is an increase in crossover distortion.  So it is already as safe as it gets.  Almost like they designed it right.

You do know that doubling the output power only increases loudness 3 decibels?  The power of the amp is limited by the power supply rail voltages and the current it can provide.  50 watts is 50 watts.  By changing the feedback resistor, I think all you are doing is increasing the sensitivity a bit, not ultimately more power.  So it may increase the signal voltage at whatever input level you have, but it won't top out any higher.  Just like turning up the volume knob.  That makes it louder at the moment, but does not increase the maximum power.

Those heftier output transistors do not HAVE more power, they can HANDLE more power.  They shouldn't alter the output at all, like any output transistor, all it does is try to do what is asked of it.  Like a 50 horsepower car and a 400hp car.  Both will drive down the road at 35 mph if that is what you ask.  But the 400hp car doesn't automatically go twice as fast all the time.   Sorta...

D7 looks correct to me.  D8,9, well one of them is likely drawn backwards.  But offhand I am nor sure which.  I THINK they are there to limit reverse bias on Q9,10.

The R2 I see in your print does not go to ground, it goes to a connector, I assume to the preamp.  it is a sample from the output bus, and so I might assume it fed some indicator circuit or something?  Would need the whole schematic

phatt

Hi Enzo,

Yes I'm aware the supply is the limitation of total power output but that assumes the Transformer is already at the limit and judging by the size of the iron it's not capable of much more. But I bit the bullet today and changed the FB resistor in the Laney to 22k. The good news is that all the voltages hardly altered after the change and the Amp was run for 2 hours into a load box with no obvious thermal issues. The sound is bigger and yes more sensitive but it will get a Real gig on Sunday so that will tell me if it has been worth the effort. If not, then I can always use another transformer and up the value of the filter caps.

Might help if I explain some detail,

This amp chassis is just bolted together (see pic). The back panel which is just alloy extrusion holds All the PSU and power amp section. undo 4 bolts and the whole section can be removed.
Sooo,,  IF the existing transformer can't handle the mod then I figure a bigger transformer can be used or simply make a whole new back panel complete with a new power amp PCB. I'm a metal worker by trade so easy to make that up from scrap metal.

The Amp is actually a keyboard type unit with 3 inputs with simple T/B tone control, very basic and not much use as a stand alone guitar amp. For guitar (both El and Acoustic) I use an out-board preamp but when I play keyboard it just plugs in the front. Modern synth keyboards don't need extra EQ so it's an ideal little rig which works well and covers my needs for the small venues I frequent.


I'm not sure I follow you on the bias pot?

The bias trim pot is wired as a 5k variable resistor linked to base and emitter of Q4 which measures ~750 Ohm when set to idle at around ~20mV.
Now as you increase that pot much past 1k Ohm the Amp will start to pull more current, limiting lamp brightness showing that to be so. With Trim pot at half way the lamp is shining bright and transformer hum becomes evident.
I'm thinking maybe the pot only needs to be 2k which might be better, or just strap a 4k7 Resistor across the pot to halve the value.

I'll keep you posted as to the outcome, and thanks for the help. :tu:

Oh I added the PDF for LaneyKB80, yes that 100k resistor (R2) does go back to common at the header plug, clearly drawn top right header plug.
Phil.

teemuk

#3
QuoteOh I added the PDF for LaneyKB80, yes that 100k resistor (R2) does go back to common at the header plug, clearly drawn top right header plug.

Those Laney printed circuit boards are quite "modular": Same power amp boards are used in a wide range of amps, and if you take a more careful look even the "pin order" of that connector is retained throughout a wide range of Laney amps.

Probably makes experimenting in R&D process much easier when you can simply in "plug&play -style" combine almost all preamp boards to almost all power amp boards, within some rational limits of course.

Some Laney amps have a limiter circuit in the preamp that is in some form driven by an attenuated speaker output signal. If the amp has a limiter the power amp board connections support the feature (you know, that "modularity" thing), if it doesn't the reduntant wire is simply used to "double" common connection for little additional bit of reliability.

teemuk

Oh, and....

Refer to this diagram concerning the correct diode orientation.

phatt

 Thankyou Temmu, that explains a lot. :tu:

The 100k resistor is indeed grounded at the preamp board plug.
For such a basic model this does not come with all the fancy stuff which is a good thing as it makes life easier for people like me who are still learning.

I got to play on the weekend but only with my El-Acoustic guitar but a bigger louder gig next Sunday where the keyboards will get a big run. <3)
Keyboards have more bandwidth and obviously will show up the limitations of the power supply.

I do have a massive do-nut transformer which delivers 35VDC rails with obviously more current ability but it won't fit without major changes. Bum  :grr  The only other transformers I have are much higher voltage and that would most likely mean a whole redesign of the power amp section as the Laney design is very basic and from what I understand may not work at higher rail voltages past ~ 40Volts.

Thanks for the picture of the diodes,, I'd say one diode needs to be flipped. I've not touched the diodes in the laney yet as I figured it wiser to only make one change at one time. So will see how the 22kFB change works for a while then I'll flip the diode D7.

I'm not worried if the Laney modification does not work out as I'll just make a whole new back section.
I've got a small collection of power devices that can be used.
i.e. 4 x 2sc5200 (250V-15A-150W) and matching 2sa1943.
I have a pair of MJ15003/4 but there is a chance they might be fakes. :-X
Oh decisions decisions. xP
Phil.

phatt

Ok I'm back to the bench for some testing,,,
after many simulations and actual test circuits I've come to realize something that I've always known but never really compared. Higher VA rating on transformer is helpful but higher voltage rails is just as important for headroom.

The question then becomes;
How high can you push these rather simple power amp designs like my Laney KD circuit when it comes to higher voltage rails?

I have 4 medium to large transformers to work with.
1/ Largest is 80V ct. gives 55 VDC rails. scary!! xP
2/ Next is a 70V ct. 48VDC rails.
3/Then a big donut monster 50V ct. gives 35VDC rails, heaps of current but likely no better in headroom than the Laney.
4/ One more giving only 32VDC rails.  (hardly worth it)

I'm posting a circuit I have up on the test bed running right now, Much like ESP P03 but using higher voltage devices and I'm getting good stable results.
Using Transformer (2) 48 V rails with a 100 Watt lamp limit, stable. :tu:

Lamp drops rails to 46 Volts at idle, turned up rails drop to around 35VDC.

Output Tr's are fine, only warm but those pre-drivers worry me as they obviously thermally cycle faster than output dev's
I notice ESP mounts the bias sense transistor right on the predrive, hum?
Sims reveal 500mW in Q5 and Q6 which does not seem like much for a 6A transistor but it gets fairly warm. Actually runs cooler when the lamp sags the rails.

I'm using 2SC5200/2SA1943 outputs (Q7,8,9,10) (250V-15A-150W)

TIP41C/42C predrives. (Q5,6)  (100V-6A-65W)

TIP41C VA (Q4)

A BD139 for bias sense. (Q3)

BC556B for input. (Q1,2)

I'm reluctant to bypass the 100 Watt lamp until I do some more research. I'm guessing the predrives and bias should be on the heat sink before I let this thing go full power.

So am I asking for a meltdown?  :lmao: Any advice most welcome. <3)

Phil.

teemuk

QuoteHow high can you push these rather simple power amp designs like my Laney KD circuit when it comes to higher voltage rails?

Depends.

Your power transistors will have some maximum collector-emitter voltage limit, and they they will have a certain "safe operating area" or SOA, which is dependent on many things, such as rail voltage and load.

Keeping the output devices within their "SOA", for instance, becomes much harder with reactive loads because of different phase in voltage and current, which might lead to much higher dissipation requirements than those estimated to purely resistive loads.

Since these "simple" design employ either a very simple scheme to force output devices to stay within their "SOA" (simply schemes simply can't do this reliably) - or no such scheme at all (even worse) - I wouldn't have high hopes that they fair decently in reliability department as soon as you change from using purely resistive dummy loads to reactive speaker loads. And those chances to survive would pretty much decrease along with increases in rail voltage.

teemuk

#8
QuoteI notice ESP mounts the bias sense transistor right on the predrive, hum?

The output is configured as "Sziklai pairs": Collector of the driver drives base.

Unlike in "Darlington pair" (emitter of the driver drives base), where we need thermal tracking of the output transistors, in Sziklai pair we must thermally track the driver transistors.

Similarly, it is beneficial to thermally isolate driver and output transistors in Sziklai pair while in Darlington pair you need to thermally couple them through common heatsink.

phatt

Quote from: teemuk on July 17, 2015, 05:12:14 AM
QuoteI notice ESP mounts the bias sense transistor right on the predrive, hum?



Unlike in "Darlington pair" (emitter of the driver drives base), where we need thermal tracking of the output transistors, in Sziklai pair we must thermally track the driver transistors.

Similarly, it is beneficial to thermally isolate driver and output transistors in Sziklai pair while in Darlington pair you need to thermally couple them through common heatsink.

Arrh Huh Light bulb moment for Me. Yes thanks heaps,, now i get it. :dbtu:

Yes I guess I'm chasing rainbows but at least I tried it. xP
I read long ago that these simple designs only work up to ~40 Volt rails, past that you are on the edge.

An interesting observation came up during all this fun;
A lot of power amplifier circuits claim X Wattage but few refference this to an input voltage and as I've discovered by simulating at least 6 different circuits, one needs to read the fine print.

All these with 1 volt input;

My stock Laney KD50 =35 Watts

Laney with 22k FB (was 10k) = 44 Watts (but earlier clipping)

ESP P03 60W circuit = 41 Watts

Then an old circuit from years back called Digi 125
which I built years back can only deliver 6 Watts

A kit Amp I built claiming 50 Watts from 35Volt rails was a rather silly 5 Watts. :o

(All above running 35~37Volt rails)

Then a Ross PC5100 (old 5Ch mixer Amp) running 55 Volt rails can only produce 8.5 Watts. even with 55 Volt rails,,, :o ???

Obviously the sensitivity to input is quite different and the above Ross Amp can easy output a lot more watts but needs at least 3 Volts at input to get 100 Watts.

I actually swapped the preamp signal between the Ross Amp and the Laney power amp and the difference was obvious, the laney being much louder at the same input level. Of course the Laney would obviously hit the rails earlier if one kept dialing up the signal.
All interesting stuff and handy to know when working on gear.

Thanks For the help Teemu,, greatly appreciated.
Phil.

phatt


Well after many attempts,, I've finally got something to work that outperforms the existing power amplifier in my Laney Amp.  :)

A simple A/B (high volume) test reveals this circuit way more capable of producing clean signal at high level, rails read 47VDC at idle and only dropping to 44 Volts at full volume. No load voltage of the transformer after rectification is around 48 ~ 49VDC.








Looks like a pile of crap but running flat-out through a load box for many hours now. The heat sink will be a bit bigger when finally built but nothing getting so hot you can't touch it. 8|

The pre-drive Q6 does run slightly warmer than Q5 but this seems to happen to other similar designs I've tested.
Q5,Q6 actually cool down a bit at high volume,, I assume that is a common issue?

I did have some issues with biasing this as well as some oscillations which seemed to be related to each other. :loco

I was using a little 150pF ceramic at C3 which i finally realized was getting quite warm,, whoops that's not right! but after some (much) scrounging of old parts I found a high quality 250pF cap and that fixed it. The Amp was then easy to bias with no squeal and also ran much cooler so I'd guess it likely had high freq instability issues.

Yes I know I need a scope, a sig-gen and and and,, Nope can't afford it all and just more stuff to have to understand.

There may well be better devices/circuits but this is all made from junk I had lying around and has cost me next to nothing cept for a few resistors.

One Question still bothers me;
Do I need Q4, 5 and 6 on the main heat sink with power transistors?
at the moment I have Q4 tracking power transistors and it's stable while Q5,6 are floating on seperate heat sinks. (see pic)

I note the Peavey Renown amp posting has pictures of the layout and predrivers are shown floating on small island heat sinks but the output is different, all NPN devices.


I'd like to thank member *Hawk* as while following his amplifier saga, the Peavey Renown Amp Schematic gave me some clues as to how to design my circuit. The front end is close to the same.

Next step is to make a PCB so still a long way to finish line.

Phil.

phatt

#11
I'm still testing and testing and,,,, :cheesy:
I was not happy with the rather high temperature of the predrive,,,so another test bed Amp design.
The previous circuit was tricky as the bias point was quite touchy, by the time you got pristine clean the circuit was running hotter. This circuit is much more civilized.


free picture upload


photo hosting

Of course a larger heat sink helped a bit but Q6 pre-drive was still running way too hot in the previous circuit,, Now Pre-drive Q5 and Q6 run at 25 deg C and don't move while Q4 runs cooler at high volume but only by a smidgen.
The circuit has been running flat out now for over 2 hours and I can't get the temperature of Power transistors much past 35 Deg.
Day temp here was about 16Deg C

For the first test I had a PTC 185 Poly switch/fuse on the output and it finally tripped and shut down the output, so I now know the circuit must be delivering at least 60 Watts clean because they are rated at 62 watts @ 8 Ohms. (Trip current is 3.7Amps)
I've now installed a PTC 250 (112Watts @ 8 Ohms, trip 5Amps) and it has not tripped ,,yet.
Interesting point these Poly switches actually darken as they heat up and get close to trip current which gives a visual indication of the limit. handy. 8|

Of course I did an A/B swap between the original Laney power amp and the new test circuit and there is no doubt I'm out performing the Laney, both perform much the same until around halfway at which point the Laney quickly runs out of puff (Rather obvious Distortion) while the test circuit stays clean till about 3/4 dial,,, So I'm quietly confident this will be a stable circuit.
Observations,
Considering this is a rats nest of long wires which is far from ideal there does not seem to be any major instability issues which I've read can cause problems so once mounted on a pcb all should be fine. (cross fingers) :-X
I'm guessing installing the 4 x 47R base resistors on the power transistors has made a big difference to the stability of the circuit so I'm getting the itch to plan out a pcb.  <3)

Meantime if anyone can see major flaws in my circuit feel free to add comment. 8)

Bare in mind this is not for hifi so no need for all the frills of the hi end designs, just a bit more clean headroom over the original Laney power stage.
If I can do this from junked pcbs at minimum cost I'm ahead. :dbtu:

I had considered adding the mandatory current limit but I'm not sure how to calculate values and as the poly switches work so well maybe that is just adding needless complexity?
Obviously the circuit can safely deliver somewhere close to 4Amps so a 5Amp shut down might be OK.
Phil.

phatt

Opps,, forgot to change bias setting on the drawing,, it should have read 500 Ohms Not 1k2 :-[
Phil.