I recently acquired a Dared MP-15 hi fi amp, it is a pleasing looking shiny thing! Using two KT 94's and four '6N3's'.
It sounds pretty good with inefficient speakers, like my little Wharfdale diamond 4's.
Unfortunately with my Lowther EX3's in Hedlund Horns, it is too bass heavy, and too powerfull!
I have 'triode'd' it by connecting R22 direct from pin 4 to pin 3 of the KT, -lowers the output!
If I disconnect the feedback loop from the 8 ohm term. Gain increases! But no noticeable tone effect!
The circuit looks overly complicated to me! Any suggestions for modifications!
I need lower gain but a nice clean throughput!
I emailed Dared, and they were nice enough to promptly send me the schematics. Only a couple of differences on my amp!
Here is the main 'per side' circuit, on the actual board are the additions R42 and C10, which I have scrawled in!
It sounds pretty good with inefficient speakers, like my little Wharfdale diamond 4's.
Unfortunately with my Lowther EX3's in Hedlund Horns, it is too bass heavy, and too powerfull!
I have 'triode'd' it by connecting R22 direct from pin 4 to pin 3 of the KT, -lowers the output!
If I disconnect the feedback loop from the 8 ohm term. Gain increases! But no noticeable tone effect!
The circuit looks overly complicated to me! Any suggestions for modifications!
I need lower gain but a nice clean throughput!
I emailed Dared, and they were nice enough to promptly send me the schematics. Only a couple of differences on my amp!
Here is the main 'per side' circuit, on the actual board are the additions R42 and C10, which I have scrawled in!
There is nothing wrong with the amp. The amplifier is amplifying beyond the bandwidth of your speaker.
There are two major ways you can tweak the bandwidth. The first one is changing the coupling caps to a lower value. This is the most common way. It was done purposely in the radio era days to prevent speaker flopping when the tuner was tuned on/off station setting. The second way of narrowing response is to lower the capacitance value of cathode bypass capacitors. this creates a higher impedance signal ground for the lower frequencies. Sometimes total removal of the bypass capacitor is needed. Keep in mind that this effects the amplification factor, but what you might gain is better balance.
There is the feedback too, but that really effects the high frequency attenuation. tweaking this part of the circuit effects the mid band frequencies (presence control is a special adustible neg feedback)
I would reduce coupling cap to .1 uf and if it still not to your liking, remove 100 uf bypass cap.
but try both ways, (bypass cap first) and see how you like it since all you have to do it lift one side of the cathode bypass cap.
There are two major ways you can tweak the bandwidth. The first one is changing the coupling caps to a lower value. This is the most common way. It was done purposely in the radio era days to prevent speaker flopping when the tuner was tuned on/off station setting. The second way of narrowing response is to lower the capacitance value of cathode bypass capacitors. this creates a higher impedance signal ground for the lower frequencies. Sometimes total removal of the bypass capacitor is needed. Keep in mind that this effects the amplification factor, but what you might gain is better balance.
There is the feedback too, but that really effects the high frequency attenuation. tweaking this part of the circuit effects the mid band frequencies (presence control is a special adustible neg feedback)
I would reduce coupling cap to .1 uf and if it still not to your liking, remove 100 uf bypass cap.
but try both ways, (bypass cap first) and see how you like it since all you have to do it lift one side of the cathode bypass cap.
Last edited:
have you tried lifting the 100uf first? as that is more effective as no bypass capacitor would give you true linear response.
as far as the coupling caps, the first one would all you'll need to change. The coupling caps are going to give you a 5-20% change. But may induce an undesirable phase shift.
as far as the coupling caps, the first one would all you'll need to change. The coupling caps are going to give you a 5-20% change. But may induce an undesirable phase shift.
I am very familiar with the design and modifying of guitar amps, but Hi Fi is new to me, the same rules do not apply!
they do, but the don't, as guitar amps are non-linear stages by design.
the audiophile design falls in the category of the sudio mic pre amp and comp designing. The I/O in pro is balanced and the audiophile/consumer is typically unbalanced.
If you need less gain why not just turn down the volume control - that is what it does!
Converting to triode mode by moving R22 will reduce open loop gain and output impedance, but make less change to closed loop gain and output impedance because of the feedback.
You could reduce gain by reducing the 10k feedback resistor (R40?), but this might create instability problems at LF or HF.
Removing C12 will lower gain but increase output impedance. This may, or may not, match your speakers better. It would depend on the extent to which they rely on electrical damping of the bass resonance. You could get more bass!
Bear in mind that this is not a guitar amp to be 'voiced' but a sound reproducer - it is designed to output a faithful copy of the input. It won't do this perfectly with such a simple circuit, but it will try. Feedback will affect all frequencies - 'presence' controls on a guitar amp are quite different as they are specifically designed to modify the sound.
Converting to triode mode by moving R22 will reduce open loop gain and output impedance, but make less change to closed loop gain and output impedance because of the feedback.
You could reduce gain by reducing the 10k feedback resistor (R40?), but this might create instability problems at LF or HF.
Removing C12 will lower gain but increase output impedance. This may, or may not, match your speakers better. It would depend on the extent to which they rely on electrical damping of the bass resonance. You could get more bass!
Bear in mind that this is not a guitar amp to be 'voiced' but a sound reproducer - it is designed to output a faithful copy of the input. It won't do this perfectly with such a simple circuit, but it will try. Feedback will affect all frequencies - 'presence' controls on a guitar amp are quite different as they are specifically designed to modify the sound.
Thanks, a few things to try out there!
Using my Lowthers I have to have the volume pot turned so low that it is inaccurate between the channels.
I could attenuate the input just after the pot, but I would prefer that I was amplifying a larger input by a small amount rather than amplifying a small input by a large amount if you see what I mean! - Better for s/n ratio.
Hence why I would like to lower the actual gain of the amplifier circuit.
Using my Lowthers I have to have the volume pot turned so low that it is inaccurate between the channels.
I could attenuate the input just after the pot, but I would prefer that I was amplifying a larger input by a small amount rather than amplifying a small input by a large amount if you see what I mean! - Better for s/n ratio.
Hence why I would like to lower the actual gain of the amplifier circuit.
One might consider merely putting a resistive pad (attenuator - two resistors) at the input of the amp to increase the voltage required to drive it to full power?
Your Lowthers are ~100dB/1w, so you need not that much power to drive them, and not that much gain... I think that you can loose 10dB in the front end and not notice much problem in the S/N department...
Or dork the resistors on the second stage (hard to read on my screen) so that it is lower in gain. More like a unity gain stage...
Oh, you could also attenuate the signal between the first stage and the second, that might be a good idea too... easy enough. No S/N loss like that... you have too much signal coming from there now!
You can either mod the gain, or mod the level. Modding the level is easier and not likely to alter the stability any.
Check ur HF response too... see if you can pass a sinewave flat out to beyond 20kHz, or if it is falling off by 5kHz or something like that?
What other amps have you used on the Lowthers?
Lowthers rarely are accused of too much bass...
_-_-bear
But the question is why does it have too much bass?
Is it because it has too *little* highs?
A scope will tell us.
Check the square wave at 1kHz. See if it looks square on the top, and without ringing, not too over damped on the leading edge.
Post it here if you like - make sure the input square wave is square, post that too.
Your Lowthers are ~100dB/1w, so you need not that much power to drive them, and not that much gain... I think that you can loose 10dB in the front end and not notice much problem in the S/N department...
Or dork the resistors on the second stage (hard to read on my screen) so that it is lower in gain. More like a unity gain stage...
Oh, you could also attenuate the signal between the first stage and the second, that might be a good idea too... easy enough. No S/N loss like that... you have too much signal coming from there now!
You can either mod the gain, or mod the level. Modding the level is easier and not likely to alter the stability any.
Check ur HF response too... see if you can pass a sinewave flat out to beyond 20kHz, or if it is falling off by 5kHz or something like that?
What other amps have you used on the Lowthers?
Lowthers rarely are accused of too much bass...
_-_-bear
But the question is why does it have too much bass?
Is it because it has too *little* highs?
A scope will tell us.
Check the square wave at 1kHz. See if it looks square on the top, and without ringing, not too over damped on the leading edge.
Post it here if you like - make sure the input square wave is square, post that too.
This is designed to be a high gain circuit (voltage gain around 50). You could increase feedback but you might need to reduce open loop gain to compensate otherwise you may get oscillation. Reducing open loop gain is not easy, as the circuit has been designed for high gain.
One thing to try: reduce R8 to 1k, and increase R6 to 390. This will increase feedback, and reduce the gain of the first stage (although by a smaller ratio). Whether this works will depend on how close to instability the original circuit is.
If you get HF oscillation or peaking try increasing the value of C8. If you get LF oscillation or peaking try reducing the value of C10 or omitting C12.
To be honest, an input attenuator wil be simpler!
One thing to try: reduce R8 to 1k, and increase R6 to 390. This will increase feedback, and reduce the gain of the first stage (although by a smaller ratio). Whether this works will depend on how close to instability the original circuit is.
If you get HF oscillation or peaking try increasing the value of C8. If you get LF oscillation or peaking try reducing the value of C10 or omitting C12.
To be honest, an input attenuator wil be simpler!
I wonder what you think the basic problem is here? I assume that turning down the volume has not had a satisfactory effect, and this is related to the characteristics of the speaker.
It's a while since I looked, but is it not common for SETs to have a high-pass filter, often two-stage, at the input? The idea I think is to stop unmanageable low frequencies before they produce harmonics, IM, output stage distortion, or a range of transformer-related effects.
Considering this amp has plenty of feedback, then its output impedance is probably low enough to cope with a widely-varying load impedance, so wouldn't explain the perceived problem. Its frequency response is also extended by the feedback so, at least for small signals, that isn't the problem either.
For larger signals, if the load impedance falls very low, distortion can rise steeply just as the proportion of feedback left to deal with it is dwindling. The harmonics produced by the output stage are efficiently reproduced by the speakers, and peceived as bass emphasis.
Perhaps it would be good to find out what the open-loop bass response of the output stage is, and use an input filter to limit the amp thereabouts.
The idea of disconnecting the feedback and abandoning the first stage could be a useful diagnostic experiment, even if it doesn't immediately give satisfactory results.
From what I have seen of SETs and single-driver speakers, it's a delicate illusion that requires a lot of messing about to get right.
Thanks folks, and please keep up the suggestions! I shall probably try most of them out at some point. I did get this amp with a view to playing around with it's internals!
Unfortunately my old oscilloscope has breathed it's last, so I'll not be able to post any traces of my investigations as yet.
Lately I have been running the horns with an old Nad amp, which has a low volume button, very useful for the purpose. (everything packed away at the moment while I redecorate the front room though) - They sound fine using that.
Might be a little while before I have another tinker with the amp, as I slipped and fell down the stairs today, and am feeling a bit battered and sorry for myself!
Unfortunately my old oscilloscope has breathed it's last, so I'll not be able to post any traces of my investigations as yet.
Lately I have been running the horns with an old Nad amp, which has a low volume button, very useful for the purpose. (everything packed away at the moment while I redecorate the front room though) - They sound fine using that.
Might be a little while before I have another tinker with the amp, as I slipped and fell down the stairs today, and am feeling a bit battered and sorry for myself!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.