I've had some time to sit and admire the work and consider the next step. This weekend I'll be populating the front ends (all but the fets). I'm saving the semiconductor devices until the chassis is ready and they can all be installed in a well grounded enclosure. Until then, I have a decision to make that I hope I can get some opinion on.
The output power supply being very important, I considered using Plitron's LoNo transformers. The problem lies in the fact that the 1.5KVA units are nearly $600 each. I've never used Plitron before and am not really sure if the cost is warranted over transformers made by Avel which I've used in the past. Finding out can be very expensive. Anyone have any experience with these?
An idea occurred to me today. If I use 2 1KVA units per channel from Avel and stack them in such a way that their fields are 180 degrees opposed do they cancel eachother out? In other words, does the stray magnetism cancel out or does this also have an effect on the cores themselves. The 2 units from Avel have more load capacity than a single 1.5KVA from Plitron and are nearly a third the cost. The pair is also a bit smaller than the Plitron because they are not encased in a metal can. They don't have the RF grid between the windings but the PSU is filtered eight ways from sunday so I'm not even sure that matters.
The output power supply being very important, I considered using Plitron's LoNo transformers. The problem lies in the fact that the 1.5KVA units are nearly $600 each. I've never used Plitron before and am not really sure if the cost is warranted over transformers made by Avel which I've used in the past. Finding out can be very expensive. Anyone have any experience with these?
An idea occurred to me today. If I use 2 1KVA units per channel from Avel and stack them in such a way that their fields are 180 degrees opposed do they cancel eachother out? In other words, does the stray magnetism cancel out or does this also have an effect on the cores themselves. The 2 units from Avel have more load capacity than a single 1.5KVA from Plitron and are nearly a third the cost. The pair is also a bit smaller than the Plitron because they are not encased in a metal can. They don't have the RF grid between the windings but the PSU is filtered eight ways from sunday so I'm not even sure that matters.
Thank you very much Nelson. I assumed that this would only truly work if the stray fields really are equal and opposite otherwise I would just end up with a different shape and possibly magnitude field than what I'd started with. If the stray field doesn't really effect the circuit though I guess there's no good reason to buy a canned transformer that costs so much more money unless you mount externally like in a tube amp and care what it looks like.
On another note, I had someone mention something to me that I found very interesting about your First Watt. He pointed out that the theory behind the amp is very symilar to the VFD drives of modern large industrial electric motors. The drive systems are irrespective of supply voltage amplitude because at any given point, the change you are trying to impose is a change in torque which is a function of current. Because the motors are AC and have no brushes the current is a function of, voltage frequency, relative position of the commutator to the stator, and the motor's instantaneous (kind of) RPM. It sounded very similar to thw First Watt being a controller that imposes a change in voice coil acceleration via a proportional change in current not voltage. So I wondered if you considered a means of taking into account the voice coil position and frequency. They are major factors effecting the impedence of the system and therefore, I assume, current/acceleration. It may also be that I completely misunderstood the design.
On another note, I had someone mention something to me that I found very interesting about your First Watt. He pointed out that the theory behind the amp is very symilar to the VFD drives of modern large industrial electric motors. The drive systems are irrespective of supply voltage amplitude because at any given point, the change you are trying to impose is a change in torque which is a function of current. Because the motors are AC and have no brushes the current is a function of, voltage frequency, relative position of the commutator to the stator, and the motor's instantaneous (kind of) RPM. It sounded very similar to thw First Watt being a controller that imposes a change in voice coil acceleration via a proportional change in current not voltage. So I wondered if you considered a means of taking into account the voice coil position and frequency. They are major factors effecting the impedence of the system and therefore, I assume, current/acceleration. It may also be that I completely misunderstood the design.
Well, moving on with the (2) 1KVA idea I guess I'll need to start with a soft start circuit. The real power switch of the A75 is a high current thiristor. What if I place a cap fed by a resistor that slows its charge time in series with the gate of the thyristor. Would this provide a slow turn on time? Slow enough to act as a soft start?
Dinu,
On my A75 the Plitron toroid is 750 VA works well with the Triac BTA40-600B,
there is a set of 2 relays to cut the +- rails operated
with a switch on the front pannel {But at startup the lights have a real deep ...}
Look at these:
http://sweb.cz/Frantisek.Rysanek/U2008B.htm
got 3rd picture...
http://www.fm.vslib.cz/~kel/research/projects/trafostart/TrafoSTART_e.pdf
http://sound.westhost.com/project39.htm
http://www.riviera-acoustics.com/catalog/product_info.php?cPath=124&products_id=900&language=en
Regards
Alain.
On my A75 the Plitron toroid is 750 VA works well with the Triac BTA40-600B,
there is a set of 2 relays to cut the +- rails operated
with a switch on the front pannel {But at startup the lights have a real deep ...}
Look at these:
http://sweb.cz/Frantisek.Rysanek/U2008B.htm
got 3rd picture...
http://www.fm.vslib.cz/~kel/research/projects/trafostart/TrafoSTART_e.pdf
http://sound.westhost.com/project39.htm
http://www.riviera-acoustics.com/catalog/product_info.php?cPath=124&products_id=900&language=en
Regards
Alain.
Dinu,
May be...
I have 2 Plitrons 750 VA custon for the A75 {Peter Daniel sold me 2 of them}
one is working perfectly, the other is actually for spare... {I thought about putting 2 in the A75 ; but .... so much projects and so little time !}
For my last projects SE 6C33C-B and FVP5A phono,
I have used custom Sumr toroids ; and I am really happy with !!!
http://www.sumr.com/
A little cheaper than Plitron, and a real great value, custom made to your choice...
Feel free to ask Richard ; he has been very helpful.
Richard <sales@sumr.com>
Regards.
Alain.
May be...
I have 2 Plitrons 750 VA custon for the A75 {Peter Daniel sold me 2 of them}
one is working perfectly, the other is actually for spare... {I thought about putting 2 in the A75 ; but .... so much projects and so little time !}
For my last projects SE 6C33C-B and FVP5A phono,
I have used custom Sumr toroids ; and I am really happy with !!!
http://www.sumr.com/
A little cheaper than Plitron, and a real great value, custom made to your choice...
Feel free to ask Richard ; he has been very helpful.
Richard <sales@sumr.com>
Regards.
Alain.
a75 boards
Hi Dinu: If you'll email me at daylne52@gmail.com I'll give you my shipping address for the boards
Hi Dinu: If you'll email me at daylne52@gmail.com I'll give you my shipping address for the boards
hallo, here are pictures of my A75... http://www.majerovi.com/richard/A75/A75.htm
nevertheless due to not stabil DC offset it was already reworked and it is now different amplifier with different sound...
nevertheless due to not stabil DC offset it was already reworked and it is now different amplifier with different sound...
That Is a beautiful amp Richard. I've been having the same problems with dc offset. It was mostly due to a 3 position switch I had made to allow me feedback selection by varying R81 value. Also, there's little room on the boards for a decent size heat sink on the 220 devices. The operating points would change very drastically with tempereture but eventually stabilize. (it was a clearly audible change which may also have been due to the output stage warming up). I really liked the feedback swith though. It does make a difference in the sound. I found that badly recorded rock (Metallica, some of my buddy's German punk, etc.) sounded grating without feedback. The sound was way too bright. It seemed as though higher frequencies were way overcompensated for listening on regular home stereos at high volume levels where you'd get alot of distortion and normally starts to sound muddy. With 100% feedback, the sound softened a bit. Live music, well recorded vocals, and good jazz sound spectacular with no feedback. At first they seemed too bright. It was just my poor point of reference. The more I listened, the more I came to the conclusion that it was much more a natural sound than I had become acustomed to from a replay system. Well, the first amp was for my friend and it now lives in Florida. I was there over Thanks Giving and had a chance to listen again. I LOVE THE A 75. It kills me to listen to my Acurus Integrated.
On a totally separate note, I was thinking of seeing what happens when I drive the A75 through a pair of coupling transformers from Jensen. ANY OPINIONS/SUGGESTIONS? My preamp is a BZLS with a nice step attenuator I got from Dantymax. 128 steps of resolution and balanced output for about 70 bucks, if I remember correctly. I highly recommend it. It’s compact and works great.
Alain. I looked at the different options for a soft start. Thanks. It looks like Elliot's version is, by far, the most workable. I can get 50W resistors from PCX for only 5 bucks each but they don't really fit his board. Also, I was thinking of incorporating a soft start and an output protection circuit all in one. If I make a dedicated 12V supply, I may as well add the extra relays for the output and be done with it. SUGGESTIONS ARE ALWAYS APPRECIATED.
On a totally separate note, I was thinking of seeing what happens when I drive the A75 through a pair of coupling transformers from Jensen. ANY OPINIONS/SUGGESTIONS? My preamp is a BZLS with a nice step attenuator I got from Dantymax. 128 steps of resolution and balanced output for about 70 bucks, if I remember correctly. I highly recommend it. It’s compact and works great.
Alain. I looked at the different options for a soft start. Thanks. It looks like Elliot's version is, by far, the most workable. I can get 50W resistors from PCX for only 5 bucks each but they don't really fit his board. Also, I was thinking of incorporating a soft start and an output protection circuit all in one. If I make a dedicated 12V supply, I may as well add the extra relays for the output and be done with it. SUGGESTIONS ARE ALWAYS APPRECIATED.
Thank you. Current status is completely different. I have new mechanical parts except heatsinks and here http://petrh.org/web/images/phd75/phd75_web.pdf you can see current schematic.
I will add additional pictures soon. A man on the picture helped me with redesign.
I will add additional pictures soon. A man on the picture helped me with redesign.
Richard, this looks very different from NP's original A75. Can you explain some of the design changes? The first thing I noticed is that you are no longer driving a balanced input. Why?
How do you adjust bias? I guess it's fixed?
How does this design take care of your drifting DC?
What are all the extra transistors for?
For your heatsinks try this guy out.
http://cgi.ebay.com/HEATSINK-ALUMIN...ryZ31489QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I've bought my set from him. He does a nice job and the price is pretty great. I Couldn't even get close to his prices from any place else, especialy for the size I needed. The new amp will have (4) 8-3/4" pieces of this section per channel.
How do you adjust bias? I guess it's fixed?
How does this design take care of your drifting DC?
What are all the extra transistors for?
For your heatsinks try this guy out.
http://cgi.ebay.com/HEATSINK-ALUMIN...ryZ31489QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I've bought my set from him. He does a nice job and the price is pretty great. I Couldn't even get close to his prices from any place else, especialy for the size I needed. The new amp will have (4) 8-3/4" pieces of this section per channel.
I meant heatsink on not power side (PCB transistors). My one looks pretty the same as you can see on the pictures. I don't use symmetrical inputs. No source available and when I had one (Denon DCD3000, it was anyway not used).
regarding schematic would be only case to ask my friend who helped me. I'm not deeply involved about this function (not experienced enough)
regarding schematic would be only case to ask my friend who helped me. I'm not deeply involved about this function (not experienced enough)
I'm having a hard time finding a suitable soft start for my 2KVA application because of the very high current requiremnts. I did find a very nice delayed turn on relay from Digikey
http://rocky.digikey.com/WebLib/Crydom/Web Data/DSD, DLD Series.pdf
I'm thinking using a large balast resistor in series in series with a thyristor switch. After the 0.1s delay, the relay will short across th resistor and voila. The only problem is that I need a DC supply for the relay.
http://rocky.digikey.com/WebLib/Crydom/Web Data/DSD, DLD Series.pdf
I'm thinking using a large balast resistor in series in series with a thyristor switch. After the 0.1s delay, the relay will short across th resistor and voila. The only problem is that I need a DC supply for the relay.
sweet boards, thank you. The A75 is a bit more complex for someone like me, but i've always been interested in it. Anyways I have a few questions, excuse the noobness of them. JP1 turns on/off feedback I'm assuming? What is the purpose of IC 1,3,4...feedback related? Do I keep the front end rails at +/-77 volts if my output stage consists of 6 irfp240 and 6 Irfp9240, with my output stage in this configuration, what should the output rails be at? Again the boards look great, thanks!!!
I have link for you... but pages are not existing in English... in case of any interest, write me and I will help you... http://www.ezk.cz/stavebnice_moduly/xsx2102.htm
http://www.ezk.cz/stavebnice_moduly/xmx1152.htm
Price is around 10-15$
http://www.ezk.cz/stavebnice_moduly/xmx1152.htm
Price is around 10-15$
MEGA,
Check out NP's original article. You can build the amp exactly the same as power values are concerned. The jumper is not for feedback, it's for DC servo. It turns DC servo on and off. It consists of the three IC's (opamp and 2 +/- regulators) plus related voltage reference zeners, caps and resistors. You may not need it if you make R81 a fixed value. Your feedback comes from R81's value. 100%=short, 50%=500R, and 0%=open. You'll notice an extra pad at R81. That's so you can place a 3 pos switch to R81 and make feedback selectable. The third pad is for shield if you choose to use a shielded connection to the switch. You may also want to make the jumper an external switch in anticipation of any effect it may have on sound. If you make R81 selectable don’t bother with a switch for the jumper. Adjust DC offset as best you can with R81 set to 50% feedback and then activale the DC servo. Without it, the offset runs away when you turn the switch.
There's alot of discussion about how many output pairs to have. It all depends on how big a PSU you have available and how much heat sinking you can provide. I made the first one with 6N's and 6P's and it works great. My next one will have a ridunkulous amount of head room hence 12 pairs. I changed those 10K resistors to ground to 20K because I am using 64V output rails. If you are using 42V rails as originally spec'd you can change these to 10K again. There should be about 10-15V more on the front end than the output PSU. The front end needs to swing full output signal + bias.
Also, check out the trim pots. Great idea from Bob Ellis. They are 25 Turn cermets. They will allow much finer adjustment when setting bias and offset.
RICHARD
Thanks. I guess this is a soft start. My problem is capacity. I'm turning on 2KVA and about 272,000uF per monoblock. I need a very high current, high power unit. Those ramp up devices from Crydom are pricy but definately suit my needs. What are these good for?
Check out NP's original article. You can build the amp exactly the same as power values are concerned. The jumper is not for feedback, it's for DC servo. It turns DC servo on and off. It consists of the three IC's (opamp and 2 +/- regulators) plus related voltage reference zeners, caps and resistors. You may not need it if you make R81 a fixed value. Your feedback comes from R81's value. 100%=short, 50%=500R, and 0%=open. You'll notice an extra pad at R81. That's so you can place a 3 pos switch to R81 and make feedback selectable. The third pad is for shield if you choose to use a shielded connection to the switch. You may also want to make the jumper an external switch in anticipation of any effect it may have on sound. If you make R81 selectable don’t bother with a switch for the jumper. Adjust DC offset as best you can with R81 set to 50% feedback and then activale the DC servo. Without it, the offset runs away when you turn the switch.
There's alot of discussion about how many output pairs to have. It all depends on how big a PSU you have available and how much heat sinking you can provide. I made the first one with 6N's and 6P's and it works great. My next one will have a ridunkulous amount of head room hence 12 pairs. I changed those 10K resistors to ground to 20K because I am using 64V output rails. If you are using 42V rails as originally spec'd you can change these to 10K again. There should be about 10-15V more on the front end than the output PSU. The front end needs to swing full output signal + bias.
Also, check out the trim pots. Great idea from Bob Ellis. They are 25 Turn cermets. They will allow much finer adjustment when setting bias and offset.
RICHARD
Thanks. I guess this is a soft start. My problem is capacity. I'm turning on 2KVA and about 272,000uF per monoblock. I need a very high current, high power unit. Those ramp up devices from Crydom are pricy but definately suit my needs. What are these good for?
Check out Peranders' site for a soft start circuit that is scalable to various sizes by varying the resistors. You might want to consider cascading two such circuits, perhaps with three delays - Power on transformer 1, bypass resistors for 1, power on xformer2, bypass resistors for transformer 2. This would keep the surge down to less light dimming levels.
Edit: Mega - I wouldn't go much fewer than 12 pairs at Dinu's elevated rail voltages, especially into low impedance/reactive loads. 6 pairs per channel will work well at the standard 38V rails. Read up on Safe Operating Area to understand how many pairs you need.
Edit: Mega - I wouldn't go much fewer than 12 pairs at Dinu's elevated rail voltages, especially into low impedance/reactive loads. 6 pairs per channel will work well at the standard 38V rails. Read up on Safe Operating Area to understand how many pairs you need.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.