I made a pair of Zaph Audio ZRT's with scan speak drivers and I'm looking to make an amp of equal quality. The simple SE seems like something I can handle.
My initial thoughts are to go with some KT88 and James OPT. Are there other OPT out there that I should consider? Elctra-Print, Tamura, etc.
Also what about power transformer and choke? Are there brands to look at that would mate well with KT88's and James OPT.
Lastly would upgrading the caps or resistors make a performance difference? Something like Black Gate.
This forum has been a wealth of information and I'm looking forward to starting this project.
Thanks
My initial thoughts are to go with some KT88 and James OPT. Are there other OPT out there that I should consider? Elctra-Print, Tamura, etc.
Also what about power transformer and choke? Are there brands to look at that would mate well with KT88's and James OPT.
Lastly would upgrading the caps or resistors make a performance difference? Something like Black Gate.
This forum has been a wealth of information and I'm looking forward to starting this project.
Thanks
For high end try Hashimoto, they have all the iron, power, choke, output. Try Vcap TFTF for coupling. For bypass caps, try Elna similic II.
http://www.tube-amps.net/Hashimoto_Products.htm
http://www.tube-amps.net/Hashimoto_Products.htm
[shill mode]If you wanted to save yourself time and a good chunk of change you could just look here...
http://www.diyaudio.com/forums/showthread.php?p=1742758
[/shill]
Of course half the fun is building it, right? The trafo recommendations made so far are good ones, I'd toss Sowter into the mix as well. There's no shortage of choices but sourcing them, lead times, and shipping can and will play a role for sure. I'd also suggest that you take care with the power supply aspect, I think it matters every bit as much as the output iron that you use. There's a myriad of choices for passive components as well, I'm a Holco guy mostly for tube gear and that's worked well for me so far.
For KT88 in the Simple SE, I'd expect about 450 volt B+ supply. Use a 510 ohm cathode resistor, and I'd recommend an output transformer designed for 3.5K~5K loading on the primary. Be sure to consider the nominal impedance of your speakers, and confirm the transformer you select has a compatible secondary winding. Take a look at George's transformer page.
If "high end" is your design criteria, I expect there will be recommendations for transformers from Hasimoto, Sophia Electric, Lundahl, and Silk. You could probably do well with transformers from James, Electra-Print or Magnequest as well.
I've started a Wiki page with contact information for various transformer suppliers. Please take a look at it if you get the chance.
If "high end" is your design criteria, I expect there will be recommendations for transformers from Hasimoto, Sophia Electric, Lundahl, and Silk. You could probably do well with transformers from James, Electra-Print or Magnequest as well.
I've started a Wiki page with contact information for various transformer suppliers. Please take a look at it if you get the chance.
That is a great price but building it is most of the fun. I guess I could take it apart and put it back together again.
This is if it was my build, I have contemplated a build such as this before for either a Simple SE or Tubelab SE.
-Electra-Print Iron all the way around or Magnequest. Magnequest power iron can be a bit hard to come by unless you watch their forum or beg it out of Mike.
-Film capacitors on the power supply side, either Solen PPE or (preferable) ASC Oil.
-Mundorf Supreme Silver and Oil coupling capacitors (or I suppose V-Caps).
-For resistors I would use PRP's most ½watt places, Riken's for grid stoppers, and Kiwame's everywhere else except the output tubes cathode, this would be a Mills MRA.
On the Simple SE I would change a few things if I where doing it:
1. I would change the front end to use LED bias, eliminating the resistor and cathode bypass capacitor.
2. I would add a small daughter board with a cascode DN2540 current source instead of the 10M45. This would offer a few advantage, at least on paper:
- It would give an increase in load impedance over the entire frequency band compared to the 10M45.
- It has significantly lower shunt capacitance compared to the 10M45.
- It would allow you to use the "mu-follower" output on the ccs offering a much lower impedance and more drive to the grid of the output tubes. This may or may not matter in the end, as the 12AT7 can probably drive the EL34/6L6/KT88 just fine, but an option.
3. I would re-size the output tubes cathode bypass capacitor, with a KT88/6550, something in the neighborhood of 330uF with a 510R cathode resistor. This would allow you to use a high quality capacitor without selling a kidney. You could try too replacing the whole R-C with a current source, all up to you. At 330uF your -3 point should be at about 10Hz, which should be below the roll off of the output transformer.
4. I would solder everything to the PCB, no terminal blocks. Also I would jumper triode connection and no cathode feedback right at the PCB. This is just me, but I trust a good solder joint over a terminal block just about any day... maybe just old school. It isn't a huge issue though, it is nice to be able to easily remove things and switch things in and out. I wouldn't use a standby switch I don't think if I was just using the 5AR4, but you still might want to include it. Just another way to extent tube life.
5. Personally I would drop the volume pot and use a line amplifier. Though I don't know why since I am working (slowly) on a Musical Machine which is integrated. Something to me just feels right about dropping the pot on this amplifier, but it is nice to get one more active component out of the signal path. If I did this I think I would look to a 12B4A (maybe choke loaded, sitting atop a CCS...maybe), or a 6SN7 actively loaded, common cathode/cathode follower, think Extended Foreplay or VTV with CCS's. Lastly, maybe a Electra-Print Ultrapath line amplifier, all film caps, with 2 x 2C22's.
I would use EL34's but that is just me. Either way I would look for good quality NOS tubes. 12AT7's are fairly cheap to come by as NOS, 5AR4's are a bit pricier, KT88's or real Tung-Sol 6550's will set you back a bit. I would go through and if it isn't much trouble, put a resistor inline with one of the heater wires from the transformer to the pcb. Size it to drop the heater voltage to around 6.2v or so. You shouldn't hear the difference. This, along with an operating point within the limits of the tubes, and you will get nice long happy lives out of NOS tubes. In some cases many times longer than new production tubes from China or Russia.
Maybe a bit much, but you asked, hehe.
Cheers and happy listening!
James
-Electra-Print Iron all the way around or Magnequest. Magnequest power iron can be a bit hard to come by unless you watch their forum or beg it out of Mike.
-Film capacitors on the power supply side, either Solen PPE or (preferable) ASC Oil.
-Mundorf Supreme Silver and Oil coupling capacitors (or I suppose V-Caps).
-For resistors I would use PRP's most ½watt places, Riken's for grid stoppers, and Kiwame's everywhere else except the output tubes cathode, this would be a Mills MRA.
On the Simple SE I would change a few things if I where doing it:
1. I would change the front end to use LED bias, eliminating the resistor and cathode bypass capacitor.
2. I would add a small daughter board with a cascode DN2540 current source instead of the 10M45. This would offer a few advantage, at least on paper:
- It would give an increase in load impedance over the entire frequency band compared to the 10M45.
- It has significantly lower shunt capacitance compared to the 10M45.
- It would allow you to use the "mu-follower" output on the ccs offering a much lower impedance and more drive to the grid of the output tubes. This may or may not matter in the end, as the 12AT7 can probably drive the EL34/6L6/KT88 just fine, but an option.
3. I would re-size the output tubes cathode bypass capacitor, with a KT88/6550, something in the neighborhood of 330uF with a 510R cathode resistor. This would allow you to use a high quality capacitor without selling a kidney. You could try too replacing the whole R-C with a current source, all up to you. At 330uF your -3 point should be at about 10Hz, which should be below the roll off of the output transformer.
4. I would solder everything to the PCB, no terminal blocks. Also I would jumper triode connection and no cathode feedback right at the PCB. This is just me, but I trust a good solder joint over a terminal block just about any day... maybe just old school. It isn't a huge issue though, it is nice to be able to easily remove things and switch things in and out. I wouldn't use a standby switch I don't think if I was just using the 5AR4, but you still might want to include it. Just another way to extent tube life.
5. Personally I would drop the volume pot and use a line amplifier. Though I don't know why since I am working (slowly) on a Musical Machine which is integrated. Something to me just feels right about dropping the pot on this amplifier, but it is nice to get one more active component out of the signal path. If I did this I think I would look to a 12B4A (maybe choke loaded, sitting atop a CCS...maybe), or a 6SN7 actively loaded, common cathode/cathode follower, think Extended Foreplay or VTV with CCS's. Lastly, maybe a Electra-Print Ultrapath line amplifier, all film caps, with 2 x 2C22's.
I would use EL34's but that is just me. Either way I would look for good quality NOS tubes. 12AT7's are fairly cheap to come by as NOS, 5AR4's are a bit pricier, KT88's or real Tung-Sol 6550's will set you back a bit. I would go through and if it isn't much trouble, put a resistor inline with one of the heater wires from the transformer to the pcb. Size it to drop the heater voltage to around 6.2v or so. You shouldn't hear the difference. This, along with an operating point within the limits of the tubes, and you will get nice long happy lives out of NOS tubes. In some cases many times longer than new production tubes from China or Russia.
Maybe a bit much, but you asked, hehe.
Cheers and happy listening!
James
Last edited:
Do you plan on using solid state diodes or a tube rectifier? You have to decide this before you can select a power transformer. I'd go solid state because I think you can get a "stiffer" and more solid power supply but many people just like the idea of all tubes. If you go with a tube your first filter cap is constrained to about 50 or 100 uF by the tube's data sheet. SS diodes can handle much higher in-rush current.
I don't think a power transformer can color the sound. It can "bottle neck" the sound if undersized. I think you get the best deal by going with a plain old, built-like-a-tank Hammond 300 series PT but buying one rated about 150% of the required current.
Chokes are the same way, If you have a budget of X dollars you can buy either a bigger Hammond or a smaller boutique brand. Go with the one that has the greater weight and higher current rating.
This is exactly what I've been looking for. This gives me an idea of the relative expense of each.
Thanks
Frankly I'm still in the learning stages and I don't know the pros/cons of tube vs. solid state. My gut tells me I'd like to go with tubes but I don't know at this point.
djn04,
You are quite welcome, no need for compensation
. There are lots of choices out there and everyone has their own way of building and their own dream setup. If this is your first amplifier (I am only assuming, and yes I know what happens when one assumes), especially tube amplifier, then I know what it can be like facing all the choices. Just thought I would present you with my version of the "if it was me" post.
I do have to agree with what ChrisA posted (as I was typing) about iron, beefy is better. I have a soft spot in my heart for those old vintage bricks of power iron that felt like they where filled with lead. Having to small of a power transformer is like having to small of a heart as an endurance athlete. It will be stressed all the time and one day, usually without warning, fail. I tend to over design, use ½ watts even if ¼ will work, use 100v capacitors even if they will only see 60v. If your total power draw is going to be 140mA don't use a 150mA power transformer. I would use at least a 200mA power transformer, preferring a 225mA to 250mA rated model, if money and space allows. Even the use of film instead of electrolytic capacitors is a bit of over engineering. However consider this; an ASC oil filled film is rated for around 100,000 hours (give or take) within it's design ratings. They are made mostly for use as motor start capacitors. More than likely you will not be pushing them anywhere near their design ratings, meaning more than 100,000 hours of operation. On the other hand, high temp electrolytic capacitors are often rated for around 2,000 to 3,000 hours. Plus they are leaky and dry up over time. There are some places where you don't have a choice, but this is just something to consider.
All this comes down to this, over design and provide a happy environment for your tubes, and they will be the only limiting factor as far as performance and life (sorry, nothing you can do about tube life except try to extend it). Not having to worry about passives failing, values drifting all over the place, or voltages bouncing around, gives you more time to worry about other things, like which record to spin now.
Cheers
James
You are quite welcome, no need for compensation
. There are lots of choices out there and everyone has their own way of building and their own dream setup. If this is your first amplifier (I am only assuming, and yes I know what happens when one assumes), especially tube amplifier, then I know what it can be like facing all the choices. Just thought I would present you with my version of the "if it was me" post.I do have to agree with what ChrisA posted (as I was typing) about iron, beefy is better. I have a soft spot in my heart for those old vintage bricks of power iron that felt like they where filled with lead. Having to small of a power transformer is like having to small of a heart as an endurance athlete. It will be stressed all the time and one day, usually without warning, fail. I tend to over design, use ½ watts even if ¼ will work, use 100v capacitors even if they will only see 60v. If your total power draw is going to be 140mA don't use a 150mA power transformer. I would use at least a 200mA power transformer, preferring a 225mA to 250mA rated model, if money and space allows. Even the use of film instead of electrolytic capacitors is a bit of over engineering. However consider this; an ASC oil filled film is rated for around 100,000 hours (give or take) within it's design ratings. They are made mostly for use as motor start capacitors. More than likely you will not be pushing them anywhere near their design ratings, meaning more than 100,000 hours of operation. On the other hand, high temp electrolytic capacitors are often rated for around 2,000 to 3,000 hours. Plus they are leaky and dry up over time. There are some places where you don't have a choice, but this is just something to consider.
All this comes down to this, over design and provide a happy environment for your tubes, and they will be the only limiting factor as far as performance and life (sorry, nothing you can do about tube life except try to extend it). Not having to worry about passives failing, values drifting all over the place, or voltages bouncing around, gives you more time to worry about other things, like which record to spin now.
Cheers
James
a tube rectifier has a bit more built-in resistance and will have a slightly lower B+ voltage. The resistance also causes what musicians (using tube rectified guitar amps) call "sag" which is a subtle kind of compression effect that you get with an under spec'd power supply. Hifi peole might call this sound "smoother". The solid state diodes in an over spec'd power supply have a more accurate sound that some might call "agressive".
But there is overlap. You can place a 200R resistor in series with the diodes to create the "sag" effect in a solid state amp and you can "right size" the tube rectifier supply to get better dynamic range and bass with tubes. After saying all that, none of this matters if you keep the volume knob below 1/2 way like most people do.
So,.. if you know you want a tube recifier you will likely buy a transformers with slightly higher voltage and a 5V secondary. You can use a very slightly smaller and cheaper PT with SS diodes.
So I've been looking at transformers. Would this Hashimoto power transformer work?
120V PRIMARY, 320V-280V-0-60V-280V-320V @ 275 MA, 0-2.5V-6.3V @ 5A X 2, 6.3V @ 2.5A, 5V @ 3A
It's a little under what the design calls for "700 to 750 VCT (350-0-350) to (375-0-375) it should be rated for at least 175 mA"
Also what about this OPT
SINGLE-ENDED 20W AT 50HZ, 5K OR 7K PRIMARY WITH SG TAP, 4, 8, 16 OHMS SECONDARY, 25 - 70 KHZ +- 2DB, INSERTION LOSS 0.3DB, PRIMARY INDUCTANCE 23H AT 70MA (5K OHMS) / 30H AT 60MA (7K OHMS), MAX DC CURRENT 100MA
Thanks
120V PRIMARY, 320V-280V-0-60V-280V-320V @ 275 MA, 0-2.5V-6.3V @ 5A X 2, 6.3V @ 2.5A, 5V @ 3A
It's a little under what the design calls for "700 to 750 VCT (350-0-350) to (375-0-375) it should be rated for at least 175 mA"
Also what about this OPT
SINGLE-ENDED 20W AT 50HZ, 5K OR 7K PRIMARY WITH SG TAP, 4, 8, 16 OHMS SECONDARY, 25 - 70 KHZ +- 2DB, INSERTION LOSS 0.3DB, PRIMARY INDUCTANCE 23H AT 70MA (5K OHMS) / 30H AT 60MA (7K OHMS), MAX DC CURRENT 100MA
Thanks
640vct seems way too low to me. 700vct should be a minimum in my opinion. I say go with the spec'd power transformer and add a nice motor run capacitor after the choke. The Hammond 374BX would be my suggestion for good results if you are willing to spend a bit more.
I used the 374BX for my Simple SE, and I'm extremely pleased with it. The only improvement might be if they had one rated for just a bit more mA on the high voltage winding. Still, no complaints here.
Let me add or clarify a few bits here.
1) LED bias is proven to work better than a resistor and a cap, provided you find the right LED. LED's were made to emit light. Many (but not all) just happen to be very good low voltage zener diodes. What is needed here is an LED with a very low dynamic impedance across the audio range. It also must have the same voltage drop that your particular 12AT7 requires for bias voltage. This will require some experimentation and a good selection of LEDs. It does work nice when you get it right.
2) The cascode DN2540 is a better CCS than the 10M45. I didn't notice any audible difference between the two, and any measurable frequency response changes are well beyond the range of the OPT. I don't claim to have golden ears and this may be a viable option for an extreme build. It is possible to use a single DN2540 in place of the 10M45 with some resistor changes. I think that the 10M45 works better though.
3) People have repeatedly criticized the bypass capacitor sizes, but they were chosen after careful measurement of a big box full of ELECTROLYTIC capacitors on an HP component analyzer. The 1000 to 1500 uF 50 or 63 volt Panasonics have a lower ESR over the audio range than other values or brands. If you are going to use a film cap, or bypass the electrolytic with another cap, feel free to use whatever value you like, but I would think 330 uF total would be about the mimimum if CFB is to be considered due to phase shifts in the feedback path.
4) Yes, the terminal blocks are there as a convinience, especially for first time builders. Many first time builders may take their amp apart and reassemble it several times before getting the cabinet into its ultimate shape. Repeated soldering and unsoldering to the PCB can cause damage.
Many "audio extremists" claim that you should strive to eliminate as many elecrto-mechanical connections as possible. My experience in the automotive world leads me to agree. I have seen the screws loosen up over time when you set the amp on top of one of the speaker cabinets. The terminal blocks have not proven to be an issue sound wise but they can be omitted if the user desires. I do NOT recommend the standby switch. This used to be standard in many amplifiers, but with many modern transformer (Hammond) the inductive spike created from turning off the switch (opening the circuit while current is flowing) has zapped the diodes or rectifier tube in the power supply. Snubbers across the switch will help, but not guarantee a complete cure.
5) Volume pot or line stage? That depends on your preference. You are going to put a volume control in the chain somewhere.
There is one minor mistake here that has caused a lot of confusion in the past. There are motor start caps, and there are motor run caps. You never want to use a start cap in a tube amp. They are designed to provide a phase shift at 50 or 60 Hz for a few seconds to start a large AC electric motor, then they are switched out of the circuit. They are generally electrolytic or cheap film caps that are not designed for continuous duty. A motor run cap is designed to go in series with one winding on an AC motor to provide a phase shift during continuous running of an AC motor. It is designed to connect directly to the AC power line, absorb whatever transients that come down the line and pass a continuous current in the AMPS for hours, days or weeks at a time. For it to do this without exploding, it needs all of the same characteristics that we desire in a power supply filter cap. Motor run caps are rated in uF and an AC voltage. They will work fine on DC at at least 1.4 times the AC voltage (many people say 2 times). I would recomment a 370 VAC motor RUN cap for the power supply in the Simple SE. I used a 100uF 370 VAC ASC motor run cap in my Simple SE.
For an ultimate Simple SE you probably want to use a tube rectifier. The SS option will boost the B+ slightly for a bit more power, but most die hard tube heads claim that a tube rectifier sounds better. There are a few technical reasons where this could be true. It does provide a softer start up too.
The voltage is a bit low, but it will work. This transformer will not allow the amplifier to make full power output. I would look for a transformer in the specified range. The OPT's look good on paper, but Hashimoto is out of my price range, so I have never used them.
Many users have reported good results with the Hammond 374BX. I have always used the cheap Allied transformer which is made by Hammond and is somewhere between a 274BX and a 374BX, but some users don't like it. I got a 372HX for a Simple P-P and it is a very nice transformer, and it doesn't get near as hot as the Allied. In fact it only gets warm.
At least one builder used a One Electron BFT-1B power transformer. It is slightly above the recommended voltage range, but is probably OK with KT88's. EL34's are a bit more sensitive to over voltage on their screen grids.
One thing to consider is what some industries call the 80% rule. When I worked for Texas Instruments (Industrial Control Division) we required de-rating ALL components to 80% manufacturer's rating (Power, voltage, current etc). The reason for this was reliability. We designed for and got 10 years plus life with systems running 24/7/365. Many of those systems are still running after 20+ years.
Filter caps (like the EPCOS B43503 series) in switching power supplies would be subjected to high ripple current, but if you only used them at 80% of the rated voltage and ripple current you could expect to get 100,000hrs operation without problems.
Aside from voltage stress, thermal stress should be considered. By operating at 80% rating, this was seldom an issue, although we still did thermal maps and testing.
By using 1/2W resistors in place of 1/4W ones you also address the thermal issues. I prefer to stay below 70% of the resistor power rating for reliability reasons.
Ever seen an old electronic system (stereo, radio, TV, etc) with discolored resistors. You can be sure they have shifted values due to thermal stress.
Filter caps (like the EPCOS B43503 series) in switching power supplies would be subjected to high ripple current, but if you only used them at 80% of the rated voltage and ripple current you could expect to get 100,000hrs operation without problems.
Aside from voltage stress, thermal stress should be considered. By operating at 80% rating, this was seldom an issue, although we still did thermal maps and testing.
By using 1/2W resistors in place of 1/4W ones you also address the thermal issues. I prefer to stay below 70% of the resistor power rating for reliability reasons.
Ever seen an old electronic system (stereo, radio, TV, etc) with discolored resistors. You can be sure they have shifted values due to thermal stress.
You have selected a power transformer based on the brand name even if the specs are wrong. The other way around would be better. Just get a PT that matches the specs. Go for a few more milliamps if you want to be conservative and run cooler.
The power supply voltage determines the operaing point for the output tubes
It's hard to know if the specs for the OPT are good. For example 25-70K +-2DB.
I think what you'd like to know the response within the range you can hear. Edcor specs theirs as "20-20K < 1db" which is better within the range that matters. The OPT you selected has no data for the 20-20K range it could be better or worse, we don't know.
The other thing about OPTs is that you really need to know the power rating at the lowest frequency. The OPT is spec'd for 25Hz but we are given 20W at 50Hz. We would want to know how many watts at 25Hz.
An important spec for an OPT s "weight". this is a crude but still good indicator of overall sound. It is a good way to roughly compare two OPTs as in general bigger is better. It would be good to know it this is a 5 pound or 12 pound OPT.
I've been shopping for transformers too. But I'm working the other way around and once I find the transformers I like I will choose the amp design that is best for them. In my opinion the transformers are by far the most important and expensive part of the amp. So way not adjust the rest of the design to suit the iron? Using this method I like the Hammond 3xx series transformers and the Edcor CXSE25 transformers
Last edited:
They are the most expensive component, and the OPT will have by far the biggest influence on the sound of the amp.
There are "experts" that will strongly dissagree with that statement (especially the low DCR crowd) but within reason, I agree with it. A power transformer that is too small will get too hot, which will hurt its reliability and the reliability of the surrounding parts. Some power transformers can emit an audible hum or buzz, which may be OK in some applications, but not in an amplifier. Some may have coupling issues between the individual windings which can transfer interference or noise from one winding to another. This is one reason that many purists prefer a tube rectifier. SS diodes create switching noise that can be coupled back through the power transformer to the filament windings, and then into low level tubes.
These are good (but not "high end") choices. I had a Simple SE with the cheap Allied power transformer and a pair of CXSE25-8-5K OPT's. It sounded good, too good, so I decided to remove the CXSE25's and use them in a 300B amp. The test amp sounded stunningly good (with some top secret 307A tubes). That was over a year ago and that amp still isn't finished.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.