• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Hammond 1650n and Dynaco mark III circuit.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey All,

I was looking for a circuit to build a pair of mono blocks around the Hammond 1650n OPT. I was thinking of using the Dynaco Mark III. Hammond says they are capable of 60 watts and can be paired with KT 88's. But I have read in some threads that they may be to lightweight for the MARK III circuit. Triode's replacement for the Mark III weighs 12 pounds while the Hammond weighs 8. This would lead me to believe what I have read may be correct. So could the circuit be used with more conservatively run-and therefore long lasting-KT 88's. or a different topology entirely. And to use KT 88's would you just give them less bias voltage to decrease their output?

Kevin
 
well, the power ability varies with bandwidth . the below sample is 60watts from 20 to 20k, and 120watts from 30hz to 15khz. so you can limit power by limiting bandwidth (input coupling capacitor)

mark III sample
Transformer specs are: 4300 ohms CT primary with screen taps, 4,8 and 16 ohm secondary. 20 Hz to 20 KHz response at 60W, within 1 db, 30 Hz to 15 KHz at 120 watts. Max DC per side (suggested) 120 ma. Exact same dimensions as the original A431 Dynaco transformer: Height 4 3/4 inches (120 mm) Width 4 inches (101 mm), Depth over covers 4 1/2 inches (114 mm), mounting centers are 3 inches (75 mm) by 3 1/2 inches (88 mm).
 
The MK3 uses global NFB. The O/P trafo has to have magnetic headroom. Otherwise, the LF error correction signal saturates the core.

If you don't want to pay $126.95 for "DynaClone" trafos, Edcor's $92.87 model CXPP100-MS-3.3K will do very nicely.

The Hammond 1650N would be a good fit for a pair of the EL34/6CA7/KT77 group. In things Dyna, that means a ST70 stereoblock or MK4 monoblocks.
 
Thanks All,

The only reason I'm using the 1650n is that I found one used on eBay that I got for $38 and another new that I found on the same day for $69. Just like I was given a 101B amp out of an old Magnavox console. And then I found a 101A (which is nearly identical) for $22.50 on eBay. I'm cheap. I am aware of the Triode OPT but I don't want to invest that much money in these amps. I just like building them for fun.

Hammond specs the 1650n at 60 watts with a frequency response of 30 to 30k. So If I understand the saturation problem correctly, at lower power output the frequency response would be better than that.

As always Eli you explain the problem succinctly and with true clarity.

The Mark IV is rated at 40 watts from 10 to 20k. If I have this right if I build a Mark III as it approaches full output it will lose bandwidth. If I build a Mark IV it will keep the wider frequency response but at a lower output. So from a non-purist perspective it isn't a huge difference because there isn't much material below 30 cycles anyway.

The real question is this. Would the KT88 Mark III produce significantly more distortion above 40 watts or would it just lose bass performance?

Kevin
 
Kevin,

It's not mere distortion. It's a sonic catastrophe, when the O/P trafo core saturates. Build with EH 6CA7s, along MK4 lines. Work backwards from the speaker, in terms of voltage swing. To get near 60 W., employing 1650Ns and KT88s, you will need a B+ rail well over 500 V. It was not an accident that I pointed you at 3.3 KOhm primary Edcor "iron", for use in a KT88 amp.
 
It's a sonic catastrophe, when the O/P trafo core saturates.

I may offer a slightly different opinion here. First off OPT saturation sounds bad. However OPT saturation in a circuit that uses a fair amount of global negative feedback sounds really bad, and can in extreme situations damage the output tubes.

So, in your opinion you can't build an audiophile quality amp with the 1650n. Or any Hammond transformer for that matter?

I have built some really good sounding audiophile quality amps using lesser transformers than these. The key is to avoid wrapping feedback around the OPT, or use only a small amount of feedback.

If you feed the OPT a large enough signal signal at a low frequency its core will saturate. This means more signal input to the core will not produce any more signal output, only more distortion. A feedback circuit will try to correct this condition by feeding more signal to the OPT making the problem worse. The current in the output tubes goes up, making them unhappy too. If the overload were to go away instantly there is still a recovery time associated with return to normal.

The type of music that you play and the impedance characteristics of your speakers play a big role here too. If you want to thump the walls with techno or rap and your speakers have a fairly low impedance at the low end of the audio range (less than 8 ohms below 80Hz) then saturation will be a big problem. If you are using smallish woofers or a typical full range speaker the impedance goes up as you approach resonance so saturation is almost impossible. My Yamahas are 10 ohms at 120 Hz rising to 25 ohms at 80 Hz. They remain above 12 ohms until 50 Hz. I can crank 50+ WPC through 5 pound OPT's.

I have been building amps with some rather smallish OPT's and I find that the usual pentode circuit with a bunch of global feedback can be problematic. I find that running big tubes like the KT88 in triode mode can make a good sounding amp with little or no feedback necessary. See this thread for an example and a lot of information:

http://www.diyaudio.com/forums/tubes-valves/133034-6l6gc-ab2-amp.html?

Another possibility is a pentode design with local feedback. See this thread for several variations on the same design in power levels from 17 WPC to 250 WPC:

http://www.diyaudio.com/forums/tubes-valves/151206-posted-new-p-p-power-amp-design.html?
 
Thanks Eli,

I understand what the major components of a tube amp do. But is there a book you could recommend that would give me an understanding of the individual components? What I mean is the different Preamp/driver circuits? And the resistors, that is the position and how to derive the values for a new design? I know what a cathode resistor is, but what is a "snubber'? I mean, when you design a race car from scratch you start with the tires that are available. Because you can't build your own tires. I would assume with a tube amp you would start with choosing the output tube. With some research I might have fewer and more interesting questions.

thanks again, Kevin
Thanks again.
 
Hey Tubelab,
Thanks for the help. I have a pair of Dynaco Mark VI's that I rebuilt using a 12AU7 front end and KT88 outputs. They sound wonderful and play really loud without any feedback at all. And surprisingly the bass is great. I listen to mostly women musicians, English rock and jazz. I'll check out those links.
Kevin
 
Hello Everyone,

I'm liking the Mark IV option more and more. But if I may impose on you all with a couple questions.

I have a pair of power transformers that I powered up and checked last weekend. They have 397-0-397 high voltage, 6.7 volts for the heaters. I don't see a bias tap. The amps they are in use 6V6 outputs with cathode bias? Is this 397 too far off the IV's 370?

Second, Rob Hull from Tube Depot published a modification to convert from 7199 to 6U8A for the driver tube. Will there be a difference sonically? (Obviously these will not be Mark IV's so this may not be a valid question.)

And, like a 58 Ford this amp only has about 17 parts. So are the circuit boards necessary? I mean couldn't it be built using terminal strips and all point to point wiring?

And, lastly, does anyone have a suggestion on a transformer for the bias supply?

Thanks for any help you may provide, Kevin
 
I mean, when you design a race car from scratch you start with the tires that are available. Because you can't build your own tires.

OK, this analogy carries a bit more weight than you think. First off, power is power only the units are different. 1 HP = 746 watts, fact. Voltage is like RPM and current is like torque. In SAE units HP = torque * RPM / 5252. In the electronics world power in watts = volts * amps. We need to put the power to the road, but the engine spins too fast and doesn't have enough torque to drive the wheels directly so we use gears. The engine torque is multiplied by the final drive ratio and applied to the wheels. A tube circuit operates at too high a voltage and not enough current to drive a speaker directly, so we use a transformer. The transformer works exactly like a transaxle, except in most cases we have only one (gear) ratio. As with a car, if you were to change the tire size (speaker impedance) a different final drive ratio would be needed to achieve the same acceleration performance.

You have a set of output transformers, which is like starting to build your race car with a gear set. The gear set (like many) has a a maximum power rating. So what you need next is to build an engine that will put that much power to the tires given your chosen ratio. As with an engine you can make a given power level by building a little torque using 4 cylinders and spinning it to 10,000 RPM (a lot of voltage), or making a V8 that runs at 4500 (low voltage). Your transformers fall in the mid range V6 level and need about 6000 RPM (about 430 to 450 volts of B+). Now you can build a Ford, Chevy, or even a Toyota to get there (choice of circuit). Each may have advantages and disadvantages, and like engine choices, each will have its loyal followers, and detractors. I prefer Mopar myself, but have built Fords and Chevys.

And, like a 58 Ford this amp only has about 17 parts. So are the circuit boards necessary?

Do you want to buy a chassis kit (like an Art Morrison) or do you want to sit down with some Chromemolly and a welder? Rolling your own allows the most flexibility, but using a PC board means that the chassis design work has already been done and the suspension geometry is right, if the chassis (board) designer is competent. A PC board from a competent designer should have all the hum and oscillation bugs worked out so you just build it and it works. I usually advise first time builders to use a PC board, but copying a known good layout works too.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.