In post #1193 I suggested to layout a PCB that comprises both the MONGREL AX frontend and the PB250 backend.
I did the layout myself. 😀
Here it is:
The dimension of the PCB is 235 x 125 mm.
I wonder if Pete likes / approves it.
Best regards - Rudi_Ratlos
I did the layout myself. 😀
Here it is:
The dimension of the PCB is 235 x 125 mm.
I wonder if Pete likes / approves it.
Best regards - Rudi_Ratlos
Hello Pete,
Nice work you do.
Did you investigate the clipping behavior of the AX1.3? i just simulated it (as a background task at office) and see some nasty clip for 1.9V input. Seems that Q10 is saturated.
Hope this helps.
Impressive THD anyway!
regards
Laurent
Nice work you do.
Did you investigate the clipping behavior of the AX1.3? i just simulated it (as a background task at office) and see some nasty clip for 1.9V input. Seems that Q10 is saturated.
Hope this helps.
Impressive THD anyway!
regards
Laurent
Just found a quick and dirty work around : connect a diode between Q10 collector and Q9 base. Now it's ok. You can also have a smooth clip by adding a small serie resistors... that in fact only push away the hard clipping...
In post #1193 I suggested to layout a PCB that comprises both the MONGREL AX frontend and the PB250 backend.
I did the layout myself. 😀
Here it is:
The dimension of the PCB is 235 x 125 mm.
I wonder if Pete likes / approves it.
Best regards - Rudi_Ratlos
Three Questions
1 Have you had a chance to listen to your build?
2 Are you willing to share it to Laser Printer Transfer Builders?
3 What PCB layout program did you use.
Excellent work by the way, keep it up, Looks Great...😎
Working on just that ! Cordell's VAS "active shunt" solves that problem (below 1). Also, the AX1.4 will be TMC / TPC /CMC ** universal. I did saturate Q10 on my 45v setup , It did not get warm or let magic smoke out , BUT ... replacing the Q9- 220R collector resistor with Cordell's transistor (A/Q3 - pix 1) is foolproof. If more than 13ma is passed by R19 , Q3 will shunt the base of Q9.
Also , the AX made a "tick" (small thump) ... this was because the VAS CCS stabilized before the input stage one did. Solved this by tying the 2 current sources together (B- pix 1). (Below 2 ) is the universal layout for the compensations , note "LC" and the 2 "TPC" pad groups.
I was told to make this circuit 110% , so be it.
"Impressive THD" , Yes at 40V p-p ..... at 80V still good. By using a 3Q wilson mirror at the input stage , I could achieve .4PPM @ 1k - 100V p-p. I am debating whether to do this for 2 reason's, #1 - the increased impedance of the wilson and #2 - how much more perfect does this have to be ? 🙂
Another decision is to whether to leave the cascode , a pair of 2sc1845's are plenty quiet by themselves. 😀
** TMC= transitional miller compensation
TPC= two pole compensation
CMC= conventional miller compensation
Nice work, Rudi ... you might want to incorporate both "improvements" before you build.
OS
Attachments
Last edited:
1 - Yes - see below
2 - Yes
3 - I used EAGLE to layout OStripper's MONGREL
Comments:
1 - I need OStripper's "O.K." to offer my layout to the DIY community
2 - I want to incorporate the latest changes (see post of McLeod) in my layout
3 - I need a "guinea pig" to prototype my layout
I will not offer my layout until it has been proven by somebody.
I have already somebody in mind to do the prototyping.
I can't do this by myself, since I do not possess an oscillospe, a function generator, ... 😱
For example, KRISFR:
The pinout of the KSA 2992, that OStripper uses, is different from the pinout of the 2SA992 my German distributor offers.
I want to be sure that my MONGREL layout works with the components that are currently distributed.
Have some patience, please.
Best regards - Rudi_Ratlos
Greetings from Bosnia,great amp,still waiting for the right box🙂.
Psu is Apex,with protection.
Here are the pictures, sorry about bad english😛
Psu is Apex,with protection.
Here are the pictures, sorry about bad english😛
Hi Nikko, Great looking build!!! Thanks for sharing the pictures. How are you tinning the PCB? Thanks, Mark
Those ARE nice. This helps me alot.. 🙂 Most builders like the 120 watt version , this is good , SO ... the "all in one version" will be a 4 device one (The one I will sell) GOOD job !! 🙂
Did you run them ??
OS
Again, thank you, me personally not interested in power.
Amplifier is a reliable, worked with the test transformer(2x45V 500W) almost all day in full force.
Of course the artificial resistance, several resistors in series.
Since I have Fullrange speakers(sonido) a few watts is al ot for a listening.
Greeting
Amplifier is a reliable, worked with the test transformer(2x45V 500W) almost all day in full force.
Of course the artificial resistance, several resistors in series.
Since I have Fullrange speakers(sonido) a few watts is al ot for a listening.
Greeting
Hey Pete
What is maximum rail voltage for the 250 and watt would the wattage be into 8? Recommended VA rating...?
Any thoughts of a Protection circuit?
Anyone have a GOOD suggestion as to a NICE case to put Rudi's amp in as is with 8 output devices... NOT overly expensive...😀
What is maximum rail voltage for the 250 and watt would the wattage be into 8? Recommended VA rating...?
Any thoughts of a Protection circuit?
Anyone have a GOOD suggestion as to a NICE case to put Rudi's amp in as is with 8 output devices... NOT overly expensive...😀
You could go 75+ V or even 80 V as the cap multiplier drops about 2 v to the voltage board. You might want to limit input amplitude like I do as this amp can create QUITE the transient.
(below) I have finalized the PB250 / AX / EX (simple supersym) , only one board left to do (artu's full comp. amp - CLX).
OS
(below) I have finalized the PB250 / AX / EX (simple supersym) , only one board left to do (artu's full comp. amp - CLX).
OS
Attachments
250W mono or 250W+250W stereo?
a transformer of 200VA will successfully power an amplifier with between 100W and 200W of maximum output.
If you have 250W+250W of maximum output, you will find that 500VA to 1000VA will give a successful build. 45Vac or 50Vac secondaries will give power either side of the 250W target. Don't use 4ohm speakers with these higher voltage transformers.
Will there be a update to the AX PCB Layout soon? (regarding the 1.9v clipping...)
Im currently waiting for more Presensitized PC Boards from my supplier before im starting on making those AX boards. Just received 2SC2240 and 2SA970 transistors a couple days ago. Only missing those boards, and 1k trimmers(only have 2k and 200r) and C92 on PB250 board and then im ready.
Done with the PB250 boards by the way, their only missing the AX board and C92 and they are ready to go! 🙂
Im currently waiting for more Presensitized PC Boards from my supplier before im starting on making those AX boards. Just received 2SC2240 and 2SA970 transistors a couple days ago. Only missing those boards, and 1k trimmers(only have 2k and 200r) and C92 on PB250 board and then im ready.
Done with the PB250 boards by the way, their only missing the AX board and C92 and they are ready to go! 🙂
AX 1.3VBC
I still need the parts , but it should work. the only differences are the way the CCS's get their reference ground , the lack of a cascode , and the Cordell safety device (Q9). (Below) is the artwork (zipfile) and a pix of the results. I have tested these changes on the real one by "point to point" (cob jobs 😀).
Also , CMC , TMC , and TPC compensation is available , read the new Cordell book or the negative feedback thread. This amp is close to his example , but tested in the real world. I will just add this as a variant to the AX package as it uses ECB standard 120 Vce input stage devices (keep it simple).
OS
I still need the parts , but it should work. the only differences are the way the CCS's get their reference ground , the lack of a cascode , and the Cordell safety device (Q9). (Below) is the artwork (zipfile) and a pix of the results. I have tested these changes on the real one by "point to point" (cob jobs 😀).
Also , CMC , TMC , and TPC compensation is available , read the new Cordell book or the negative feedback thread. This amp is close to his example , but tested in the real world. I will just add this as a variant to the AX package as it uses ECB standard 120 Vce input stage devices (keep it simple).
OS
Attachments
Last edited:
Can you provide an updated version of the schematics as well?
Many thanks and best regards - Rudi_Ratlos
Many thanks and best regards - Rudi_Ratlos
PSRR - the next level !!
THD is important , but PSRR is just as much a distortion mechanism. It also has a second effect to give better separation .. hence , a better sound stage. Since I have been using cap multipliers , this has been quite apparent in my listening experiences. (Below 1) is the best performance/simplicity ratio (circuit wise) I have been able to produce. Some would "go the mile" and use a fully regulated supply , but by looking at attachments (Below 2 - OPS /3 - input stage) , you can see we have stellar PSRR. Even at the output , this is more than adequate. 🙂 I have only added a darlington to the multiplier (MJE340/350 - ksc1845/ksa992) , this improves the multiplier by 100X , the only downfall is a greater BE drop (2V+). OH WELL !! The price you pay for performance. 😀
OS
THD is important , but PSRR is just as much a distortion mechanism. It also has a second effect to give better separation .. hence , a better sound stage. Since I have been using cap multipliers , this has been quite apparent in my listening experiences. (Below 1) is the best performance/simplicity ratio (circuit wise) I have been able to produce. Some would "go the mile" and use a fully regulated supply , but by looking at attachments (Below 2 - OPS /3 - input stage) , you can see we have stellar PSRR. Even at the output , this is more than adequate. 🙂 I have only added a darlington to the multiplier (MJE340/350 - ksc1845/ksa992) , this improves the multiplier by 100X , the only downfall is a greater BE drop (2V+). OH WELL !! The price you pay for performance. 😀
OS
Attachments
Nice OS, thanks! 🙂
One question, the diodes Murs120, 5A on the output boards, im using MUR820 there, this should be ok, right? Their the only ones i have on hand with such speed and current handling.
Oh, and at the AX board, 1n4148 should work instead of the specified 1N914?
One question, the diodes Murs120, 5A on the output boards, im using MUR820 there, this should be ok, right? Their the only ones i have on hand with such speed and current handling.
Oh, and at the AX board, 1n4148 should work instead of the specified 1N914?
??? Murs120 = DIODE ULTRA FAST 1A 200V SMB - MURS120T3 = 1A diode.
I use either this or 1n4900 series ... "superectifier" http://www.vishay.com/docs/88509/1n4933gp.pdf
For the "big ones" ... 3A for the pb120/250 will do the job (flyback immunity) 5A is optional. I use these now ... http://www.vishay.com/docs/88744/srp300a.pdf Faster switching is somewhat important seeing what these "flybacks" do.
OS