A short update for the weekend.
The Alpha Tester has just returned home late Friday evening from business trip.
I am sure we all understand to give him time to recover and look after his family, etc.
At least you want to wait for his listening impressions before jumping into anything.
In the meantime, maybe you can consider whether you want to copy our build 1:1; or just the F5P frontend + M2OPS.
If you want an exact copy, you will need to have access to some machining.
But that applies to almost all DIY projects, not ?
Here are some information about the case we used, and a comparable one from HiFi2000.
The links are just examples, you need to search around for best offers.
BZ3213 Blank
https://www.ebay.com/itm/295542655383
https://www.aliexpress.us/item/2255800533216694.html
External Dimensions -- 320W x 130H x 313D
Internal Dimensions -- 220W x 122H x 300D
Heatsink Dimensions -- 300 x 50 x 125
Weight 6.3kg; Shipment weight 8kg
Hifi2000 Mini Dissipante 3U 300mm
https://modushop.biz/site/index.php?route=product/product&path=285&product_id=716
External Dimensions -- 330W x 126H? x 313D?
Internal Dimensions -- 250W x 120H x 300D
Heatsink Dimensions -- 300 x 40 x 120
Weight: 5.5 kg
The BZ3213 has slightly better heatsinks.
The Mini Dissipante has a bit more width inside.
We have not tried the latter, but we attach here a layout drawing of our build.
You can then check whether you can still use all our PCBs.
Or you have to plan your own build from scratch, using other cases.
You need a DWG reader to view the file; plenty of freeware around.
Patrick
.
The Alpha Tester has just returned home late Friday evening from business trip.
I am sure we all understand to give him time to recover and look after his family, etc.
At least you want to wait for his listening impressions before jumping into anything.
In the meantime, maybe you can consider whether you want to copy our build 1:1; or just the F5P frontend + M2OPS.
If you want an exact copy, you will need to have access to some machining.
But that applies to almost all DIY projects, not ?
Here are some information about the case we used, and a comparable one from HiFi2000.
The links are just examples, you need to search around for best offers.
BZ3213 Blank
https://www.ebay.com/itm/295542655383
https://www.aliexpress.us/item/2255800533216694.html
External Dimensions -- 320W x 130H x 313D
Internal Dimensions -- 220W x 122H x 300D
Heatsink Dimensions -- 300 x 50 x 125
Weight 6.3kg; Shipment weight 8kg
Hifi2000 Mini Dissipante 3U 300mm
https://modushop.biz/site/index.php?route=product/product&path=285&product_id=716
External Dimensions -- 330W x 126H? x 313D?
Internal Dimensions -- 250W x 120H x 300D
Heatsink Dimensions -- 300 x 40 x 120
Weight: 5.5 kg
The BZ3213 has slightly better heatsinks.
The Mini Dissipante has a bit more width inside.
We have not tried the latter, but we attach here a layout drawing of our build.
You can then check whether you can still use all our PCBs.
Or you have to plan your own build from scratch, using other cases.
You need a DWG reader to view the file; plenty of freeware around.
Patrick
.
Hi all,
Yes, I am the lucky alpha tester – this project has been through development for a while, it took a little time to gather everything that was needed, test each part and so on to make sure everything worked together. “Perfectly Integrated” means F5P voltage stage and M2OPS together, but I think we wanted more than that – all items had to work nicely together. Naturally this meant some building and testing.
My usual approach when I build an amp is to test one channel, then the other, make basic adjustments, and then once it is in the ballpark of bias, offset etc, I will let it play for several hours in the workshop. All went well in the workshop so after (almost) final adjustements I moved it into the listening room. The system is comprised of various turntables, cartridges and phonostages for analogue, all of high enough standard to reveal a lot, and different builds of Miros excellent AD1862 dac with SD card player, USB and Bluetooth streaming. My normal amps cycle through Aleph and F5 builds, various valve amps, and one or two AB amps as well. The most used one (and the one I used for comparison in this initial audition) was a combo of the Iron pre and a AB amp using similar topology to F5 front end with Exicon lateral fets for the output stage. Speakers are a freshly serviced set of stacked Quad 57s. Perhaps I am just neurotic, but I like to leave an amp running for 30-60mins before doing any critical listening – this lets everything come to a stable temperature etc. I noted that change in temp from cold to fully warmed is about 26-28degC and DC offset was stable throughout.
The very first impression was that this amp has remarkable preservation of detail – I would have thought my normal system was very resolving, and had an excellent soundstage, but the F5Pi really makes me re-evaluate that. While trying to avoid the usual jargon, maybe I can describe an example of this. You know when you are listening to something and some note or instrument sound is at the far right or left of the soundstage? Often those small little sounds are there if you are listening for them, but often don’t stand out on their own. Well, the F5Pi brings them right out into the light. Now imagine that across all the soundstage and you have this hugely detailed soundscape in front of you. I’ll freely admit that I really enjoy a big and deep soundstage, and the F5Pi is a master at this. In comparison, when I use my DHT preamps and different power amps, I get a different presentation, maybe more forward projection but narrower.
As I listened more, something else that stood out very much was the dynamic impact from the amp – things like snare hits, piano key impact etc are more than I have heard from many other amps. Another test I use are massed violins – if an amp tends towards any kind of sharpness, I always think the violin will reveal it. All was good in this regard, no harshness evident. Bass, while bearing in mind the limitations of Quad 57s, was full and deep and very tight indeed.
I also brought the amp to a friends place, and we had a small gathering there to get other peoples impressions so that it wasn’t just these ears. So with 3 other people we listened to a wide range of music, some of which would not be regarded as “well recorded”! Main amp for comparison was the J2 (a clone built with Xen devices and EUVLs published schematics). The J2 has been a favourite among this group for a while now so I was curious to see what the guys thought. They essentially confirmed my comments above – massive amounts of detail and impact but with no harshness whatsoever. As a result of this visit, I have a few invites to others so I will add any more comments that I get as I make those visits.
Impressions like this are so subjective, I know it’s really just one person’s opinions, but having listened to many DIY and commercially available systems I think my ears are reasonably well educated.
Yes, I am the lucky alpha tester – this project has been through development for a while, it took a little time to gather everything that was needed, test each part and so on to make sure everything worked together. “Perfectly Integrated” means F5P voltage stage and M2OPS together, but I think we wanted more than that – all items had to work nicely together. Naturally this meant some building and testing.
My usual approach when I build an amp is to test one channel, then the other, make basic adjustments, and then once it is in the ballpark of bias, offset etc, I will let it play for several hours in the workshop. All went well in the workshop so after (almost) final adjustements I moved it into the listening room. The system is comprised of various turntables, cartridges and phonostages for analogue, all of high enough standard to reveal a lot, and different builds of Miros excellent AD1862 dac with SD card player, USB and Bluetooth streaming. My normal amps cycle through Aleph and F5 builds, various valve amps, and one or two AB amps as well. The most used one (and the one I used for comparison in this initial audition) was a combo of the Iron pre and a AB amp using similar topology to F5 front end with Exicon lateral fets for the output stage. Speakers are a freshly serviced set of stacked Quad 57s. Perhaps I am just neurotic, but I like to leave an amp running for 30-60mins before doing any critical listening – this lets everything come to a stable temperature etc. I noted that change in temp from cold to fully warmed is about 26-28degC and DC offset was stable throughout.
The very first impression was that this amp has remarkable preservation of detail – I would have thought my normal system was very resolving, and had an excellent soundstage, but the F5Pi really makes me re-evaluate that. While trying to avoid the usual jargon, maybe I can describe an example of this. You know when you are listening to something and some note or instrument sound is at the far right or left of the soundstage? Often those small little sounds are there if you are listening for them, but often don’t stand out on their own. Well, the F5Pi brings them right out into the light. Now imagine that across all the soundstage and you have this hugely detailed soundscape in front of you. I’ll freely admit that I really enjoy a big and deep soundstage, and the F5Pi is a master at this. In comparison, when I use my DHT preamps and different power amps, I get a different presentation, maybe more forward projection but narrower.
As I listened more, something else that stood out very much was the dynamic impact from the amp – things like snare hits, piano key impact etc are more than I have heard from many other amps. Another test I use are massed violins – if an amp tends towards any kind of sharpness, I always think the violin will reveal it. All was good in this regard, no harshness evident. Bass, while bearing in mind the limitations of Quad 57s, was full and deep and very tight indeed.
I also brought the amp to a friends place, and we had a small gathering there to get other peoples impressions so that it wasn’t just these ears. So with 3 other people we listened to a wide range of music, some of which would not be regarded as “well recorded”! Main amp for comparison was the J2 (a clone built with Xen devices and EUVLs published schematics). The J2 has been a favourite among this group for a while now so I was curious to see what the guys thought. They essentially confirmed my comments above – massive amounts of detail and impact but with no harshness whatsoever. As a result of this visit, I have a few invites to others so I will add any more comments that I get as I make those visits.
Impressions like this are so subjective, I know it’s really just one person’s opinions, but having listened to many DIY and commercially available systems I think my ears are reasonably well educated.
Last edited:
I've had a few queries about how this amp compared to the J2 - what were differences etc.
I'd have to start by saying this was all done in a room with a few guys, and certainly wasn't a controlled environment. So while I believe I have captured it correctly, you need to allow for that.
The J2 mentioned above (clone based on EUVL's published sch) has been a favourite of some in our group. It has a really deep soundstage, and manages to pull this trick with vocals in particular where they are projected forward in the mix. When you match that with the correct speakers, both combine to produce something very attractive indeed. In my friends place, the J2 was partnered with a 01a based preamp (essentially Ale Moglia's gen 2) which is also no slouch in that department. So the overall impression here is that the soundstage is (front to back) deep, with forward projection of vocals, and no shortage of bass or top end sparkle. Its a really very attractive combination.
The F5Pi changes that up quite a bit. When we switched over from the J2, the initial impression is that everything is more precise - and certainly that detail I mentioned before really jumps out. Bass was more precise, which I think is what people mean when they say bass was tighter and more "grippy". The soundstage is different as well - that forward projection of vocals that the J2 has is not there, but the soundstage is larger left to right and top to bottom, and has matching depth front to back as well. Just like with the J2 combo, this is very attractive as well, but in a different way. It is just that bit more precise and clean sounding, even though there is no detection at all of any kind of noise etc from the J2. I reckon these are just different flavours depending on what you like.
Its hard to describe stuff like this without reverting to magazine review language!
I'd have to start by saying this was all done in a room with a few guys, and certainly wasn't a controlled environment. So while I believe I have captured it correctly, you need to allow for that.
The J2 mentioned above (clone based on EUVL's published sch) has been a favourite of some in our group. It has a really deep soundstage, and manages to pull this trick with vocals in particular where they are projected forward in the mix. When you match that with the correct speakers, both combine to produce something very attractive indeed. In my friends place, the J2 was partnered with a 01a based preamp (essentially Ale Moglia's gen 2) which is also no slouch in that department. So the overall impression here is that the soundstage is (front to back) deep, with forward projection of vocals, and no shortage of bass or top end sparkle. Its a really very attractive combination.
The F5Pi changes that up quite a bit. When we switched over from the J2, the initial impression is that everything is more precise - and certainly that detail I mentioned before really jumps out. Bass was more precise, which I think is what people mean when they say bass was tighter and more "grippy". The soundstage is different as well - that forward projection of vocals that the J2 has is not there, but the soundstage is larger left to right and top to bottom, and has matching depth front to back as well. Just like with the J2 combo, this is very attractive as well, but in a different way. It is just that bit more precise and clean sounding, even though there is no detection at all of any kind of noise etc from the J2. I reckon these are just different flavours depending on what you like.
Its hard to describe stuff like this without reverting to magazine review language!
Also fresh from the bench :
Just as you can use the BA3 as preamp, you can f course also use the F5P as preamp.
All the ingrdients such as source select and volume control are already there.
But the circuit needs some adaption to get down to a Gain of 10dB instead of 26dB.
Prototype works first time.
Patrick
Just as you can use the BA3 as preamp, you can f course also use the F5P as preamp.
All the ingrdients such as source select and volume control are already there.
But the circuit needs some adaption to get down to a Gain of 10dB instead of 26dB.
Prototype works first time.
Patrick
Two more frontends were built today with gain 20 using Fairchild MOSFETs and JFETs.
Bandwidth is 750kHz -3dB without any compensation or feedback caps.
So circuit works well also with the said alternative devices.
Too dark for photos now.
Tomorrow, if the sun is out.
Patrick
Bandwidth is 750kHz -3dB without any compensation or feedback caps.
So circuit works well also with the said alternative devices.
Too dark for photos now.
Tomorrow, if the sun is out.
Patrick
As promised.
Note the oritentation of the JFETs due to different pinout.
The caps shown are not connected and only come into play in MFB configuration.
Maybe should pair with some Fairchild power MOSFETs in the M2OPS.
Should have quite a bit higher 2nd harmonics.
But the frontend can still be nulled for H2 as in the F5 Turbo.
Extra pads provided for trimming resistors.
Patrick
.
Note the oritentation of the JFETs due to different pinout.
The caps shown are not connected and only come into play in MFB configuration.
Maybe should pair with some Fairchild power MOSFETs in the M2OPS.
Should have quite a bit higher 2nd harmonics.
But the frontend can still be nulled for H2 as in the F5 Turbo.
Extra pads provided for trimming resistors.
Patrick
.
Attachments
The Preamp (alone, gain of 3) will be a separate project to be published separately in 2014, once properly done.
The M2OPS is fully published in Ref.1. No secrets.
How we support those to build the integrated amp depends on how exact you want to copy.
If you are just interested in the F5P+M2OPS, we can send a BoM by post, at cost.
We can also offer complete sets of matched devices.
But to copy exactly as we have built will need a lot more documentation from Fran.
Not just BoM, but assembly instructions, trimming, measurements , ...
Will take time and effort, and when he has time.
Might consider publishing a simplified spice file for the Fairchild version.
But that also needs to wait till more measurements are done.
Patrick
The M2OPS is fully published in Ref.1. No secrets.
How we support those to build the integrated amp depends on how exact you want to copy.
If you are just interested in the F5P+M2OPS, we can send a BoM by post, at cost.
We can also offer complete sets of matched devices.
But to copy exactly as we have built will need a lot more documentation from Fran.
Not just BoM, but assembly instructions, trimming, measurements , ...
Will take time and effort, and when he has time.
Might consider publishing a simplified spice file for the Fairchild version.
But that also needs to wait till more measurements are done.
Patrick
Typing mistake.
For the preamp, first distortion measurements, then alternative source selector and switched relay volume, then new case.
But of course you can also use the Relaxied.
In that case you can start with that first.
We want to try something else because we want 6 inputs instead of 4, and a different layout.
Case will be BZ3207P, now on its way to the EU.
Patrick
For the preamp, first distortion measurements, then alternative source selector and switched relay volume, then new case.
But of course you can also use the Relaxied.
In that case you can start with that first.
We want to try something else because we want 6 inputs instead of 4, and a different layout.
Case will be BZ3207P, now on its way to the EU.
Patrick
Last edited:
To make sure everything works well in a public project :
We are looking for 1 or 2 Beta testers to try out the Public Version of the F5P+M2OPS PCBs, as well as the building instructions.
As you would surely understand, we retain the right to choose who to work with, to maximise the chance of success and minimise friction, if any.
He or she will get access to (revised) Gerber files and building instruction documents, without saying.
This only applies to the F5P+M2OPS, not the entire integrated amp.
Let's see if there are brave persons out there with the required skills and experience.
Please send me a PM in that case.
Patrick
We are looking for 1 or 2 Beta testers to try out the Public Version of the F5P+M2OPS PCBs, as well as the building instructions.
As you would surely understand, we retain the right to choose who to work with, to maximise the chance of success and minimise friction, if any.
He or she will get access to (revised) Gerber files and building instruction documents, without saying.
This only applies to the F5P+M2OPS, not the entire integrated amp.
Let's see if there are brave persons out there with the required skills and experience.
Please send me a PM in that case.
Patrick
Last edited:
I am pleased to say that we already have 2 Beta Testers, who will proceed to order PCBs.
In the meantime, Fran has sorted out his distortion analyser and here is the first measurement of the entire integrated amp.
Signal source is the Viktor oscillator at full output, attenuated by the modified Relaixed to output 8Vpp into 8R.
Pretty much the same as measured earlier with the AP2722.
We are still working on a few things to see if the 50Hz and harmonics can be lowered further.
Patrick
In the meantime, Fran has sorted out his distortion analyser and here is the first measurement of the entire integrated amp.
Signal source is the Viktor oscillator at full output, attenuated by the modified Relaixed to output 8Vpp into 8R.
Pretty much the same as measured earlier with the AP2722.
We are still working on a few things to see if the 50Hz and harmonics can be lowered further.
Patrick
