LJM MX50 kit amp

Great explanation tvrgeek, thanks a lot. My only question is, what sonic difference those make?

esgigt: what is the great difference between this amp, and let's say an Audio Analoge Puccini? The AA has an NE5532 driving TIP darlingtons. Nothing fancy or complicated, and it's a well regarded - we can say high end - amp. Or you meant some million dollar exotics by high end?

You got it right, the million dollar exotics that are not this good. But some are. Some are very very good. Heck, I sure hope my "Reasonable" amp I am working on will be a lot better. I am very curious if it sounds better than my old Creek that has such a high reputation but I find not very inspiring. I sold all my small Rotels, so I can't compare it to something like an RB 960. (The Rotel did have two output pairs) It is going into my garage for background music anyway. I have an Aimir integrated amp to use as the chassis, heat sinks and transformer. Anything will be better than it is. It was "well regarded" and I can assure you, the only good thing about it is the very nice heat sinks.

A more valid test would be to compare it to a gain-clone on the same power supply. Probably about the same cost to build if you did not use a board for the big National chip.

All I was doing in my effort was to apply what I have learned on paper to the same basic design and parts as supplied. It still is a CFP output. With only one output pairs, it falls flat below 4 Ohms. ( Yes Virginia, those 98 Ohm speakers can actually give very low impedance due to phase and back EMF. ) Most of these e-bay kits are sold to the market that thinks they are getting something for nothing. In this case the Music Fidelity MV-50. They pretty much copied and then cheapened the original intellectual property including any design choices that may not have been the best.

How does it sound compared? Well, as I am not going to build it stock, but only as I have modified, I don't know. My test will be if my wife finds it acceptable.
 
Please note there are two versions, the MX50 and the MX50X2, these are not identical topology, and not just because of the output transistors, the previous stages are also different.

I encourage you to build it, it's a cheap and nice project :)

They did two things to make it cheap. Dropped to one pair of outputs, which does make a dollar to two difference in cost, and I believe the original had a Darlington VAS. That is a 5 cent change that IMHO was a mistake. I was under the impression the basic topology was the same.

One should not forget the major cost of an amp is the power supply and chassis. A good transformer could be $100. Heat sinks almost as much. Chassis and all it takes to build an amp, a bunch more. Even if it was the original stolen intellectual property, you would have to make a very large investment to match the Music Fidelity amp. Top quality parts, careful matching, and so on. This is a cheap project because I have a cheap chassis that is good enough. I hope no where near as good as a real MF as that would be disappointing.

My next project, the "Reasonable" is a bit more advanced. I intend to built it on a higher quality chassis and I expect it to be very good for the small low power BJT amp it will be. I expect it to better than the MX-50, my Creek and maybe my modified DH-120. Will it beat my spare HCA 750? I would wager a case of fine suds not. I can give you a list a mile long of folks, many haunting these very forums, that know more about amps in their little finger than I do with a pile of books and a few years as a lab rat in industry many years ago.
 
Here are my SPICE models. Please note I could not find models for all the parts as supplied, so I made fair guesses to some. Not all the parts I used are in the LTSpice library, so you have to hunt them down as I did. Some don;t matter, like using the 1N4001 on the output instead of a 4007. The outputs I modeled are quite good and probably have higher gain.

I am going to use the original parts with the exceptions of the resistors I changed, power supply mods, and additional caps for the compensation. I have everything in my parts box. I hope I have enough stuff to make a DC sense and power up/down circuit. I don't intent to spend any cash on this. Saving that for my Reasonable.
 

Attachments

  • MX-50.asc
    10 KB · Views: 619
  • MX-50-improved.asc
    13.3 KB · Views: 528
Here are my SPICE models. Please note I could not find models for all the parts as supplied, so I made fair guesses to some. Not all the parts I used are in the LTSpice library, so you have to hunt them down as I did. Some don;t matter, like using the 1N4001 on the output instead of a 4007. The outputs I modeled are quite good and probably have higher gain.

I am going to use the original parts with the exceptions of the resistors I changed, power supply mods, and additional caps for the compensation. I have everything in my parts box. I hope I have enough stuff to make a DC sense and power up/down circuit. I don't intent to spend any cash on this. Saving that for my Reasonable.

Thanks for posting those, great contribution. I will post you some pics from LJMs own board where he dissects the MF amp. Does not look anything fancy or exotic inside, better rather cheap. I did not hear that, or have any schematic so there is nothing I can tell for sure, just what I saw. My point is so called high end is most of the time based on money, faith, misinformation and ignorance. If you build a krell your own with their production parts, same topology, same layout diy pcb, and show it to your friends, it will be some nice sounding cheapo homebrew amp. If the same thing is in a krell chassis and comes with an invoice of thousands of dollars, it will be a miracle, that presents music in an unimaginable way...
 
Here are my SPICE models. Please note I could not find models for all the parts as supplied, so I made fair guesses to some. Not all the parts I used are in the LTSpice library, so you have to hunt them down as I did. Some don;t matter, like using the 1N4001 on the output instead of a 4007. The outputs I modeled are quite good and probably have higher gain.

I am going to use the original parts with the exceptions of the resistors I changed, power supply mods, and additional caps for the compensation. I have everything in my parts box. I hope I have enough stuff to make a DC sense and power up/down circuit. I don't intent to spend any cash on this. Saving that for my Reasonable.
Thanks for sharing your models! :cool:
In the mean time I retrieved the missing semiconductor models for Spice.
The 2SD669 is rather simple compared to the others, so it will not be an exact representation.

*SRC=2SD649;2SD649;BJTs NPN;Si;SANYO 120V 1.5A 140Mhz
.MODEL 2SD649 NPN (IS=56.5F NF=1
+ BF=220 VAF=197 IKF=0.6 ISE=32.4p NE=2
+ BR=4 NR=1 VAR=20 IKR=0.9 RE=0.479 RB=1.92 RC=0.192 XTB=1.5
+ CJE=202p VJE=1.1 MJE=0.5 CJC=65.1p
+ VJC=0.3 MJC=0.3 TF=1.14N TR=789N)

.model 2SD669 NPN (IS=5p NF=1 BF=250 ISE=5p NE=1.5 IKF=3 VAF=150 RB=1 RC=0.25 RE=0.25 TF=1.14ns CJC=50p Vceo=120 Icrating=1500m mfg=Hitachi)

.model BC556B PNP(IS=3.83E-14 NF=1.008 ISE=1.22E-14 NE=1.528 BF=344.4 IKF=0.08039 VAF=21.11 NR=1.005 ISC=2.85E-13 NC=1.28 BR=14.84 IKR=0.047 VAR=32.02 RB=1 IRB=1.00E-06 RBM=1 RE=0.6202 RC=0.5713 XTB=0 EG=1.11 XTI=3 CJE=1.23E-11 VJE=0.6106 MJE=0.378 TF=5.60E-10 XTF=3.414 VTF=5.23 ITF=0.1483 PTF=0 CJC=1.08E-11 VJC=0.1022 MJC=0.3563 XCJC=0.6288 TR=1.00E-32 CJS=0 VJS=0.75 MJS=0.333 FC=0.8027 Vceo=65 Icrating=100m mfg=Philips)

.model BC546B NPN(IS=2.39E-14 NF=1.008 ISE=3.55E-15 NE=1.541 BF=294.3 IKF=0.1357 VAF=63.2 NR=1.004 ISC=6.27E-14 NC=1.243 BR=7.946 IKR=0.1144 VAR=25.9 RB=1 IRB=1.00E-06 RBM=1 RE=0.4683 RC=0.85 XTB=0 EG=1.11 XTI=3 CJE=1.36E-11 VJE=0.65 MJE=0.3279 TF=4.39E-10 XTF=120 VTF=2.643 ITF=0.7495 PTF=0 CJC=3.73E-12 VJC=0.3997 MJC=0.2955 XCJC=0.6193 TR=1.00E-32 CJS=0 VJS=0.75 MJS=0.333 FC=0.9579 Vceo=65 Icrating=100m mfg=Philips)
 
Many models are on Central Semi and OnSemi. Everyone gets the model libraries from Bob Cordel. The outputs and drivers I had to use something close as I could not find anything.

I have been adding my models to the master libraries but now I find I run the risk of LT overwritting them. Catch-22. Incorporate them they are in the menu. Add them as an .include and they are safe, but you need to know what to include. Less pick and test.

You will find models from vendor to vendor do not always match. Some more complex, some more optimistic.

Looking inside a chassis can be deceiving. What makes good ground and power layout may not look exotic. It may look messy. Fancy binding posts don't sound any better. How well they matched the IPS or output transistors is not viable. I don't know if they do or not, just you can't know by looking.
 
I've been following this thread with some interest. A couple of pairs of the MX50-SE would work great for multi-amp active loudspeakers that I like to build. But I have a question about the load:

I have some linear power supplies that I can put together that will produce unloaded +/-46VDC rails. I see that the amp is rated for 8 ohms. How about into 4 ohm loads - this is pretty common for home audio drivers these days. Can I use the amp at these rail voltages into 4 ohms or am I pushing or exceeding SOA, etc.? If it can be used into 4 ohm loads but with lower rail voltages, what would you suggest? Sorry, I am not into designing amplifiers myself so I am not sure on these things. If this amp is not suitable for 4 ohm loads, can you suggest something of class AB that is known to have good performance and does not require setting bias? Thanks!
 
That would be a very good application for these. As much as I pick on it, it is better than most you could buy.

You will have seen my comment on how it falls flat at 2. Not great at 4. Why should we worry about 2 Ohms when our 8 Ohm speakers have a DC resistance of 6 Ohms? Because they are a complex load and because of back EMF. Those 8 Ohm speakers could give you an instantaneous load of 3 easy.

All is not all is lost or the big box store would only sell Krells. If going into an active speaker, DIY has some big advantages.

Yes, you could most certainly exceed SOA unless you take precautions.

*You can add a baker clamp in the VAS and flying diode clamps on the bias servo to keep your maximum current within the SOA. I'll work these out and post them. ( three cheap diodes) I would really worry about SOA with the amount of power you are planning. I would say to do the long tedious effort to plot it, but that is not a beginner task. I am not confident in my efforts there yet.

*A 46V supply is a LOT of power for a single output pair. (11 Amps into 4 Ohms, or 500 WATTS) I think 20W is about all one pair should do, the 40 I am designing to is marginal. ( 36V supply). Just because it is there, you don't have to turn it up and use it. Call that "music power" or better, call it headroom.

*Your bigger supplies will give you better dynamic overhead.

*You could add a second output pair, but CFP outputs with two pairs may be difficult to stabilize.

*Limit the BW of the woofer section which is where the lower impedance would show up. The distortion may increase at 20K, but you don't care. Driving the tweeter ( just a blocking cap and the rising impedance of the tweeter, is actually very easy. I doubt you will ever clip the tweeter as it will be set 6 or more dB below the woofer anyway. Any distortion from the woofer amp won't blow the tweeter anyway.

* Every amp requires you to set the bias. Every amp requires you to verify the stability that works on paper works in the real world. Simulations assume perfect parts.

* If I were putting together some bookshelf active speakers, I might be trying class D or a gain clone. I suspect the MX-50 to be superior.

Coffee is ready. I'll get back to you. Snowed in today.
 
Got me thinking. You have a better idea for what I should use these for. I'll stuff a gain-clone in my Aimor chassis and use these to build an active speaker for my wife's electric piano. A couple of 12 V "laptop" supplies, and bit of vector board for the crossover & eq.

First light coming. I have the "duty" snow blower for the block so I had better get going.
 
That would be a very good application for these. As much as I pick on it, it is better than most you could buy.

You will have seen my comment on how it falls flat at 2. Not great at 4. Why should we worry about 2 Ohms when our 8 Ohm speakers have a DC resistance of 6 Ohms? Because they are a complex load and because of back EMF. Those 8 Ohm speakers could give you an instantaneous load of 3 easy.

All is not all is lost or the big box store would only sell Krells. If going into an active speaker, DIY has some big advantages.

Yes, you could most certainly exceed SOA unless you take precautions.

*You can add a baker clamp in the VAS and flying diode clamps on the bias servo to keep your maximum current within the SOA. I'll work these out and post them. ( three cheap diodes) I would really worry about SOA with the amount of power you are planning. I would say to do the long tedious effort to plot it, but that is not a beginner task. I am not confident in my efforts there yet.

*A 46V supply is a LOT of power for a single output pair. (11 Amps into 4 Ohms, or 500 WATTS) I think 20W is about all one pair should do, the 40 I am designing to is marginal. ( 36V supply). Just because it is there, you don't have to turn it up and use it. Call that "music power" or better, call it headroom.

*Your bigger supplies will give you better dynamic overhead.

*You could add a second output pair, but CFP outputs with two pairs may be difficult to stabilize.

*Limit the BW of the woofer section which is where the lower impedance would show up. The distortion may increase at 20K, but you don't care. Driving the tweeter ( just a blocking cap and the rising impedance of the tweeter, is actually very easy. I doubt you will ever clip the tweeter as it will be set 6 or more dB below the woofer anyway. Any distortion from the woofer amp won't blow the tweeter anyway.

* Every amp requires you to set the bias. Every amp requires you to verify the stability that works on paper works in the real world. Simulations assume perfect parts.

* If I were putting together some bookshelf active speakers, I might be trying class D or a gain clone. I suspect the MX-50 to be superior.

Coffee is ready. I'll get back to you. Snowed in today.

I could use lower rail voltages.

Can someone chime in about what rails would be suitable for a 4-ohm nominal load (amp connected to driver, no passive crossover) using the MX50-SE?

I will probably use the 46VDC rail power supplies with LJM L15D-Pro amp boards instead of with the MX50.
 
NO NO NO. 4 Ohms is a BAD speaker. Seriously. If there is anything I have learned while studying amplifiers is what we have been doing wrong in our speakers. 4 ohms means you are asking this little undersized 60 W amp to be a 120W amp! Run, do not walk to the nearest exit.

Oh, when you say no passive crossover, I hope you remember to always have a DC blocking cap on the tweeter.

I would not even be as stupid as to hook up a 4 Ohm nominal speaker to the 12V version. What would be safe? 6V maybe. And you would have other compensation problems.
 
What about the MX50X2 with 4 ohm? Anybody have this amp? I ordered the MX59se for a gutted receiver chassis.

4 Ohms is a bad speaker. No amp is is a good choice for 4 Ohms. You need to be stable at least an Ohm below the DC resistance. A typical 4 Ohm speaker has a Re of about 3 Ohms. Maybe 3.2. That means you need to drive TWO OHMS. 4 Ohm drivers are intended to be used in series.

If you have to drive a bad design like that, you need more output pairs. I don't know much about class D. Maybe it would work better.

Can't say about the various iterations of the MX-50. The one I got ( and posted) was not like the advertised design. Two pairs of outputs would be a big help. Yo will likely need to change the compensation values for such a bad load.