Discussion in 'DIY' started by Alan0354, Mar 29, 2017.
I'm going to assume those fly back diodes are scotkeys and that they fuse is after the cap bank
Thanks for the reply
No, the diode is normal diode, not schottky. I can easily change to schottky if needed.
Yes, the fuse is AFTER the big capacitor bank. If not, there will be a fire!!! As I said, I only have 100uF after the fuse, which is very small.
I am thinking whether it is necessary to have schottky diode for protection. I think normal diode is good enough. The regular diode turns full on about 1V. The point of using a diode is to prevent reverse bias the base emitter diodes of the output transistor. Base emitter diode of transistors behaves like a ZENER diode when reverse biased, it will break down about 3 to 6V. Once the base emitter diode start to conduct, you risk burning the transistor.
But the base emitter diode does not conduct until you have at least 3V or so, a normal diode will clamp at 1.2V maximum, so this will protect the transistors already.
So how does it sound ? What speakers are you useing ? And what element is your system ?
The THD measurement was posted in post #26 in page 2 of this thread.
My impression of the sound was posted in post #38 in page 2 of this thread.
I only have an Acurus 3X200W and Nakamichi Stasis PA-7 to compare to. Don't expect the sound characteristic to be very different between amps, nothing like changing the speakers. As I said, my amp has more separation, better sound stage, more clarity. But the PA-7 is slightly warmer and slightly smoother. My believe is my amp has very low distortion, so it reproduces everything like the source. If the original sound is raw, it will not cover it by smoothing it out.
My main speaker is a pair of JM Lab Spectral 913.1 three way floor standing with two 8"woofer each. There's a picture in the same post #38 showing the speaker.
The little one sat on top of the JM Lab is Monitor Audio. It was used as insurance when I first power up the amp in case something gone wrong and burn the speaker.
Now I realize that I'm being hypercritical but I never liked the stasises I mean it had tons of current and was very clean but sanatry too clean often jap AMPs sound busy cluttered the exception is a big Denon or stax my favorite speaker is the Duntec 1000 I did a series of mesurments and it did a really good one k square wave on well recorded grand piano with a matched pair Scropes condenses
In to a Apex Atr . 100 deck was truly real sounding
I don't know your definition of clean of the Stasis, in my definition, the Stasis is less clean as separation is not as good. To me, my amp is cleaner, more separation, more layer, that you can hear more details. That's my definition of clean.
It's hard to describe sound in words. That's why I kept saying I can only promise the THD graph, the low distortion. i cannot promise anyone that they will like the sound. If anyone like the warm sweet sound like a tube amp, I don't think my amp is it. But if you want to hear separation, the small little sound, more 3D and bigger sound stage, I think my amp is better than the Stasis.
I listened to Dave Grusin/Lee Ritenour in the GRP all-stars Live from the Record Plant, I can definitely hear better separation of all the instruments, bigger, and more in your face like you are in the middle.
BUT when I listen to Ricardo Monteneer that has a higher voice, the Nakamichi Stasis definitely gives him a smoother voice.
As for me, I use the amp to watch tv most of the time, for home theater, I definitely think my amp is better because you can hear all the little sound with more separation and clarity. For me, I would never go back to the Stasis. The Nakamichi is sitting in the garage right now. The Nakamichi is going to serve as a reference for me to beat. I am working on the design of my next amp, I am either working on error correction and/or tube/SS hybrid or something. I don't know.
Are you designing your own amp also? I like to hear more from you.
Hey Kjello and Dan
I run into a problem, the two transistors 2SA1930 and 2SC4883 I use are obsoleted lately and I cannot find anyone have them in stock. Mouser is on back order on 2SA1930A http://www.mouser.com/Search/Produc...)virtualkey65560000virtualkey757-2SA1930(Q,M)
I have been searching with no luck. I ordered 2SA1930 from Mouser already. You guys can help me finding these transistors also. I checked Digikey, Mouser, Allied and Newark. Utsource have them, but they are surplus store and I don't know whether I can trust them. These are 200MHz TO220 150V/180V transistors, I cannot find equivalent.
I have enough for myself for my own use for a long time, but you both need them.
I am looking into some slower transistors, but that will take quite a bit of time as I need to qualify the new transistors to make sure it is absolutely stable in the amp. I use 200MHz transistors so it is fast enough to get out of the way of stability. The next best I can find is only 70MHz.
These are Japanese Toshiba units?
I have a friend in Japan who might help out
Yes, they are Toshiba. Sanken supposed to make 2SC4883, but it has lead time and need large quantity. I have some 2SC4793 NPN that can still be used, I change to 2SC4883. I can spare two each of those for each of you, I might be able to salvage from the old boards also. I ordered more of the 2SA1930, but I cannot spare any out until I receive the new ones.
BTW Footfungus suggested Quest Components, they have the 2SA1930:http://www.questcomp.com/questdetails.aspx?pn=2SA1930&mpid=44312082&pt=4
I don't know how good they are, but you can research a little.
So for one amp its needed 18 X 2SA1930 and 18 X 2SC4883 ?
How many do I need to proper match pairs for 2 amps ?
No, these are predrivers, one for each board.......... two for each amp. That's the reason I rather find them instead requalify a new inferior transistor. I have about 25 of A1930 and more C4883, it's enough to last my whole life time, but not enough to share if I can help it. I am thinking about buying 100 of the A1930, those are dirt cheap transistors, just need to find them.
Concentrate on 2SC4883, when I get the A1930, I just give them to you two.
Ha ha, if it's the big transistor that I use 18 ea amp, I would not think twice to re qualify another one. That two are easy, many different sources and those two are available.
I am thinking out loud, The big transistors I use are TTC5200 and TTA1943, Mouser sell them at about $1.80 for quantity of 100. I still have enough for 3 boards. I am thinking whether you and Dan want to go in with me to buy like 100 to get the good price. You two pay me and I'll send along with the boards.
I need to confirm with Mouser, the description of the TTC5200 is correct, but the picture is wrong. It's should be TO-264 package.
Let me know. If so, I should be able to match it up for you. But no guaranty it match perfect, I got to 3mV matching on Vbe in my amp. I hope I can get as good. I have a fixture for this.
For Dan, it's simple as he live in the state, but I don't know your custom in your country.
BTW, I have not heard from Dan for a while!!!
Yes, I took a little holiday to celebrate my 25 year anniversary on July 4th! Lost my independence on Independence Day! We went out of town for a week.
I'll gladly split the cost of the transistors. Let me know how much and I will send the funds to you by whatever method you prefer. Also, I have plenty of room in the chassis to move everything around and remounting the new boards and components. I'm also designing a faceplate and can either anodize and laser etch the markings or cut on a CNC milling machine. Nothing finalized yet, but it can be simple with just a logo and on/off switch or as complicated as I want.
On July 5th, the right channel on my Marantz 1200 went out again. It started out with something loose, The right channel was sometimes low and I would tap the right power amp board and the sound would go back to normal. I've been playing the amp loud since it was rebuilt, but after close to 50 years it's just wearing out. Then, on July 5th, I was listening and heard a slight crackle and the sound was gone. Measured 52 Vdc at the relay board. I have a feeling it may be the other predriver transistor since it still has the original transistor and the failure is just like what happened the last time. I haven't had time to troubleshoot because I've been doing vehicle maintenance since I got home, and it's hot, hot, hot in Pasadena!!! My motivation for repairing the amp quickly is gone since I'm more likely to wait and build another amp!
I bought the Cordell and Self books. Sorry to say they are not the most expensive books I've bought (new engineering textbooks are at least $150). I also like to buy new since I can write and mark them.
I'm all in for the new amp. Please let me know how much.
Good to hear from you Dan.
I am still working on the order of those transistors.
Want to show the progress.
This is shows both sides of the pcb I build for you two. I spent a lot of effort to compact down the IPS/VAS section and also worry about the signal integrity of the signals. I put in a lot of effort to have ground, two power and even the output all in planes, I did this with no compromise. The result is it's hard to build on the SMD part of the board. So I decided to build the SMD portion of all 4 boards for you two. I even put all the transistors on the back side. This will make it a lot easier for you to continue the rest of the building. Save you a lot of ordering parts also. You can put in all the through hole parts.
This is how the complete IPS/VAS section looks like, I yet to put in the two big caps as it is in the way. You see I put on copper take to wrap the two matching transistors to get good temperature tracking. I already picked out match pairs for you two. They match to 1mV or better. In my first amp, the output offset is less than 5mV.
This is a picture of all 8 boards I am building. The 4 on the left are two each for you two, the 4 on the right side are my own. You see I already putting in a lot of the through hole parts.
You can see clearly a FULL ground plane on the top side of the pcb ( the side with the most through hole components). You can see on the bottom side two power planes on each side of the pcb, the middle plane is the output plane. ALL DONE ON A TWO LAYER BOARD!!!
This is not an easy board to build, you see components at all different angles to achieve shortest signal path, robust continuous planes for ground, power and output. Lots of jumper wires needed as you can see some yellow wires already put in. As I said, this is my amp, I don't compromise, I want the best. The board is 2mm thick, it's strong, much harder to break due to stress.
This is not an easy board to build, I made it the best I know how for performance. I try to match the critical transistors to make it more robust in sharing the power dissipation. I spent a lot of time on this.
Actually all the boards are ready to send out already, I am just waiting to bring up my boards to confirm and do the necessary modification before sending them out. Together I will send three of the filter pcbs out for filtering. Now that I know both of you can layout the circuit exactly like in my amp, chances of successful should be much better. You can see in the pictures in this thread. The placement of boards, transformers, filter caps will be the same.
My plan is to power up my new board, make sure it is stable, run the THD, post that as then send out the boards. Then we can build the amp together and I'll show pictures on my amp as it progress.
Mathcing i guess important for smooth class A operation, and keeping heat production lowest possible..
Great, of course Ill split costs.
Customs here probably the highest known to man
Nice work man, Looks great !
Separate names with a comma.