I finally found this post again in my print out of this forum. I've looked but don't see or missed if you got a answer to your question about whether your results from testing your DB's were good. I plan to a test a MountainView DBtest now that my M2x with Tucson boards is working and sounding terrific. Last thing I want to have happen is insert a new DB and blow things up.
Thanks.
Hello after the break🙂
Some time I spent listening to the amplifier with the Ishikawa card. I can't describe music so beautifully, so in short: I'm impressed with the stereophony and the general culture of sound. I can and I want to listen loud and louder without getting tired. Cool.
About the hum, it is heard with the ear in the loudspeaker. Mumetal boxes have lowered the noise of 150hz by a few decibels and is now at the same level as 100hz and 200hz. What else can I do to get rid of this? Let me remind you, that the 100hz ripple from the power supply is at the level of 170mV
Some time I spent listening to the amplifier with the Ishikawa card. I can't describe music so beautifully, so in short: I'm impressed with the stereophony and the general culture of sound. I can and I want to listen loud and louder without getting tired. Cool.
About the hum, it is heard with the ear in the loudspeaker. Mumetal boxes have lowered the noise of 150hz by a few decibels and is now at the same level as 100hz and 200hz. What else can I do to get rid of this? Let me remind you, that the 100hz ripple from the power supply is at the level of 170mV
Initial power up and testing
So, after getting my BA2018 line stage preamp up and running I moved on to the M2X. I have it all assembled and getting ready to turn it all on. The power supply section appears to be working ok as I'm getting steady +26.3V and -26.2V with no load, but haven't connected power to the main M2X boards yet.
My question is, are there any checks/tests I can do by powering up the main boards on their own first? Or should I just install an input stage and go from there?
So, after getting my BA2018 line stage preamp up and running I moved on to the M2X. I have it all assembled and getting ready to turn it all on. The power supply section appears to be working ok as I'm getting steady +26.3V and -26.2V with no load, but haven't connected power to the main M2X boards yet.
My question is, are there any checks/tests I can do by powering up the main boards on their own first? Or should I just install an input stage and go from there?
Check and recheck all, then get a current limit and go for it.... at some point, ya have to power it up.
Keep a hand on the power... In fact, I pulse the power at first. On 123, sniff and touch for heat, do a visual. Just be careful what you touch. Anyway, I have built 10-12 amps in the last 15 months. At some point you have to do all you can for safety and then, "engage." 🙂
Verify each part and channel alone...
JT
Keep a hand on the power... In fact, I pulse the power at first. On 123, sniff and touch for heat, do a visual. Just be careful what you touch. Anyway, I have built 10-12 amps in the last 15 months. At some point you have to do all you can for safety and then, "engage." 🙂
Verify each part and channel alone...
JT
Last edited:
Mark explained very well how to do it in this post(nr. 3826)
https://www.diyaudio.com/forums/showthread.php?p=6406871
Last edited:
> A new input stage daughter card for M2x: Cedarburg / IPS8 (Jan 2021)
Mark,
I do not quite understand your motivation of using an AD797 this way, instead of the original simple source follower with no global feedback.
Is it because the JFETs have become unobtanium ?
If so, I would have thought that a CFP source follower would be a more natural successor for the original ?
(This can be done with all active parts in SMD, for example).
And if it has to be an opamp, how about an OPA627, which is much more forgiving and essentially equal in low noise / low distortion ?
Is there a sonic advantage with the AD797 ?
Or is it only to demonstrate that the AD797 can be made unity-gain stable ?
Thanks in advance.
Cheers,
Patrick
Mark,
I do not quite understand your motivation of using an AD797 this way, instead of the original simple source follower with no global feedback.
Is it because the JFETs have become unobtanium ?
If so, I would have thought that a CFP source follower would be a more natural successor for the original ?
(This can be done with all active parts in SMD, for example).
And if it has to be an opamp, how about an OPA627, which is much more forgiving and essentially equal in low noise / low distortion ?
Is there a sonic advantage with the AD797 ?
Or is it only to demonstrate that the AD797 can be made unity-gain stable ?
Thanks in advance.
Cheers,
Patrick
Mark has clearly noted that Cedarburg is an exercise to use the highly-regarded AD797 in the M2x because it is a challenge. 🙂
Reports also say it sounds pretty darn good as well. 😀
Reports also say it sounds pretty darn good as well. 😀
IPS8 / Cedarburg is a toy to play with. As are all the other IPS daughter cards. Designing it was entertaining and, I hope, building and listening to it will also be entertaining.
For Calculus-II types of people who wish to analyze the Cedarburg circuit, I'd recommend Chapter XIII in James Roberge's textbook, which MIT has uploaded as .pdf under the title "Compensation Revisited" in their Open Courseware on the web. I also recommend reviewing the concept of "Noise Gain"; In my opinion, Walt Jung's explanation in Chapter 1 of his book is the best available (pdf here). When Scott Wurcer and I discussed the Cedarburg concept, we mostly talked about Noise Gain.
Once you've got the mathematical analysis complete, and have calculated Cedarburg's Noise Gain, it may or may not occur to you that Cedarburg actually will work with an entire category of forbidden opamps, opamps that absolutely will not and cannot work in either IPS7 (schematic in post #3408) or in Tucson (schematic in post #1) input cards for the M2x. And once the penny drops, once you realize that surprising fact, you can decide whether or not to disregard my strenuous recommendation: DO NOT put any opamp except AD797 in Cedarburg. You can be contrary and try the forbidden fruit; find out whether you enjoy the resulting sonics.
For Calculus-II types of people who wish to analyze the Cedarburg circuit, I'd recommend Chapter XIII in James Roberge's textbook, which MIT has uploaded as .pdf under the title "Compensation Revisited" in their Open Courseware on the web. I also recommend reviewing the concept of "Noise Gain"; In my opinion, Walt Jung's explanation in Chapter 1 of his book is the best available (pdf here). When Scott Wurcer and I discussed the Cedarburg concept, we mostly talked about Noise Gain.
Once you've got the mathematical analysis complete, and have calculated Cedarburg's Noise Gain, it may or may not occur to you that Cedarburg actually will work with an entire category of forbidden opamps, opamps that absolutely will not and cannot work in either IPS7 (schematic in post #3408) or in Tucson (schematic in post #1) input cards for the M2x. And once the penny drops, once you realize that surprising fact, you can decide whether or not to disregard my strenuous recommendation: DO NOT put any opamp except AD797 in Cedarburg. You can be contrary and try the forbidden fruit; find out whether you enjoy the resulting sonics.
Thanks for the refs and for coming up with yet another card to try 😉... I am no math surgeon, but I am interested in the general concepts as they apply to what is ultimately coming out of the speakers. I kinda intuitively see were you may be going with the noise gain as you have unity gain, and shooting for low output impedance without introducing any noise but could you give us your take in context?
I think I asked before why was the output cap present in most opamp based cards absent here?
> IPS8 / Cedarburg is a toy to play with. Designing it was entertaining ....
Sure. DIY fun does not have to be logical. 🙂
If it were supposed to be non-global-feedback, then it should not be opamp in unity-gain (with max. NFB).
If it were the Zout of the Toshiba JFETs being too high, then a JFET-BJT CFP will be good replacement.
hybrid-buffer - calvins-audio-pages
If it has to be opamp anyway, then one might as well use it for voltage gain and save the transformer.
But fun is fun. You cannot argue against that. 😉
Cheers,
Patrick
Sure. DIY fun does not have to be logical. 🙂
If it were supposed to be non-global-feedback, then it should not be opamp in unity-gain (with max. NFB).
If it were the Zout of the Toshiba JFETs being too high, then a JFET-BJT CFP will be good replacement.
hybrid-buffer - calvins-audio-pages
If it has to be opamp anyway, then one might as well use it for voltage gain and save the transformer.
But fun is fun. You cannot argue against that. 😉
Cheers,
Patrick
That configuration you suggest is completely logical. But it wouldn't be an input card for M2x...
It's alive!
I finally plucked up the courage (after some earlier guidance) to switch on my new M2X. Initially all seemed well. No pops, cracks, bangs or dimming lights in the house. But after about 30 seconds the supply rails both dropped to zero. It turned out that the fuse I had installed was only 750mA. After replacing it all went much better. After nervously waiting for the 30 second mark to pass the supplies settled down to around +/- 22.5Vdc. I left it to warm up for almost an hour, monitoring the temperature, before setting the dc offset at the outputs. I was able to get both pretty steady around 0.03mV. I was happy with that so switched off and installed a pair of MountainView daughter boards, hooked up my recently completed BA2018 preamp and a pair of old loudspeakers.
Other than a suspect connection at one of the RCA connectors on the M2X (which will be replaced tomorrow) it sounded great. Next up will be trying the different input stages to see how they all sound. I've got Ishikawa, Austin and Tuscon assembled but Norwood is still a bag of bits in a box. The SMD components were a little intimidating, but after managing those in the BA2018 I'm game to give it a go. I tried uploading some pictures but I get as far as the 'Manage Attachments' window, and when I click on the 'Choose File' button nothing happens. I'll try again later.
Anyway, thanks to all who've contributed to making this possible. I love it
Now where's that Pearl II gone......
I finally plucked up the courage (after some earlier guidance) to switch on my new M2X. Initially all seemed well. No pops, cracks, bangs or dimming lights in the house. But after about 30 seconds the supply rails both dropped to zero. It turned out that the fuse I had installed was only 750mA. After replacing it all went much better. After nervously waiting for the 30 second mark to pass the supplies settled down to around +/- 22.5Vdc. I left it to warm up for almost an hour, monitoring the temperature, before setting the dc offset at the outputs. I was able to get both pretty steady around 0.03mV. I was happy with that so switched off and installed a pair of MountainView daughter boards, hooked up my recently completed BA2018 preamp and a pair of old loudspeakers.
Other than a suspect connection at one of the RCA connectors on the M2X (which will be replaced tomorrow) it sounded great. Next up will be trying the different input stages to see how they all sound. I've got Ishikawa, Austin and Tuscon assembled but Norwood is still a bag of bits in a box. The SMD components were a little intimidating, but after managing those in the BA2018 I'm game to give it a go. I tried uploading some pictures but I get as far as the 'Manage Attachments' window, and when I click on the 'Choose File' button nothing happens. I'll try again later.
Anyway, thanks to all who've contributed to making this possible. I love it
Now where's that Pearl II gone......
DB testing
I ended up using the M2x along with a light bulb tester for amplifiers. I think I used a 40w light bulb and looked for any glow. I remember having one DB I had to fix, the light bulb never went out fully and a noticeable glow difference from the good DB to the bad DB.
I ended up using the M2x along with a light bulb tester for amplifiers. I think I used a 40w light bulb and looked for any glow. I remember having one DB I had to fix, the light bulb never went out fully and a noticeable glow difference from the good DB to the bad DB.
Glad to hear it... at first I wrote, "grow a pair," lol but then I remembered my apprehension and revised the post. Before you get all happy with changing boards... esp the unobtainium Pass original, I would run it for a week or so, a good time to make the change is when you don't look at it every few seconds, or wonder is something burning as in a dog sniffing the air. lol
Run what you brung and give it a chance to settle. The time you have on it, nothing is needs changed yet. Get some hours on it first, then make changes... just my 2
jt
Ha Ha. Thanks JT. I wouldn't have been offended with your original posting. Throughout my career I've had my share of witnessing 'letting the smoke out of the box' incidents (yes, some were mine) so am always cautious. Although these days I tend to check and double check everything before turning the power on. Especially where higher voltages are concerned!
I tried installing all of the boards I have built up so far tonight, just to make sure they worked, and all was good. The Austin ones are currently in there and will stay for at least a month or two. Who knows, they may end up staying in there a lot longer.
