I wasn't even smart enough to mark which tube it was. I put the 4 tubes from this amp aside for future experiments. If it causes trouble again it will meet the hammer. I have about 20 of these tubes but they were all pulled from scrapped equipment. All were exposed to moisture and probably were directly rained on for years. The fact that it drew current at power on without blue glow may indicate contamination between the glass and the plastic base. I have seen this before in some of my tubes.
I have cranked the same E130L's in the same board before with good results but I didn't venture above 500 volts. Maybe they just don't like 600+ volts. The screen current wasn't right from the moment I started playing with it since the voltage at the 4.7K resistor was too low after adding the second channel, which was the one that fried. I'm thinking that one of the tubes or the sockets had a problem.
The cathode in the E130L is far bigger than the typical audio tube but not quite as big as the 6HJ5's. I dissected one with a white getter and found that it is not a typical frame grid tube. The G1 rods are fat, maybe 0.1 inch with fine grid wire wound in the usual manner. The G2 rods are fatter maybe .125 inch wound in the same way. It did look like they were aligned. Don't remember about the radiator fins (I am at work now).
I had wired octal sockets on top of the 12 pin sockets. I removed them and replaced all 4 screen resistors and one cathode resistor (I broke it). I put a 1K 2W resistor in place of the 4.7K dropping resistor in the screen regulator (cheap fuse). Upon installing the 6HJ5's and powering up I could set the bias, and watch the tube current go up as I applied drive but there was no power output. The 1K resistor started smoking. DOH. I had forgotten the plate jumpers needed for capless tubes. The 1K resistor now has 4 color bands on it, all black.
After replacing the jumpers the amp worked and I measure 20 volts across the 1 K resistor at 75WPC, but I am not so sure that it is still 1K.
All experiments so far have been done using some low cost guitar amp OPT's wired for 6600 ohms. To my amazement I set the power at 75 WPC and started dropping the frequency to find the point where the distortion hits 5%. It was 44 Hz. I will try 3300 ohms tonight.
I just noticed that the E130L datasheet specs g1 dissipation at only 0.1 Watt. I tried to find a similar spec on some sweep tubes, but no luck. The 6L6GC does give curves for AB2 operation though, and the top curve shows g1 current at 40 mA at +20V on g1, for a dissipation of 0.8 Watt (peak). Would appear that g1 current draw might best be avoided for the E130L.
The red board is RC coupled to the output tube grids. That limits grid current to brief peaks. I have slammed some real grid current into several sweep tubes without issue but haven't tried AB2 on the E130L. As you suggest maybe I shouldn't.
Other users report a rather fragile screen grid too. I have been warned not to abuse the G2 spec. I have stuffed these into a Simple SE and cranked them to 400 volts in triode but many tubes can be abused in triode mode. I was running these at 150 V on G2.
Wired the OPT for 3300 ohms. Big power increase.
Cranked the drive until 5% distortion was reached. One channel made 144 watts, the other channel made 123 watts. Distortion at 1 watt is 0.7% in both channels, distortion at 50 watts is 3% in both channels. All measurements were made with both channels driven.
Cranked the drive until 5% distortion was reached. One channel made 144 watts, the other channel made 123 watts. Distortion at 1 watt is 0.7% in both channels, distortion at 50 watts is 3% in both channels. All measurements were made with both channels driven.
Well looks like everything should be a go for me. Just waiting on those parts. I will have to see if I can find a 1k resistor for R21. I think I have a wirewound one in a box that I bought for another project. I might have to go digging. That is some serious power that is being pulled out of this board. I know you have gone much higher but this seems pretty sane and doable for a guy like me.
R21's only purpose is to absorb some of the dissipation that would otherwise be burned up in the mosfet. I think that the 4.7 K would work fine with the 6HJ5's but I don't have one handy. I had 3 15K's in parallel but took those out when I rebuilt the board. I stuck a 1K in there because its real easy to measure the screen current. Volts across the resistor = milliamps. I had a jumper in there for a while too. I am using the heat sink from a Pentium 1 chip and it barely gets warm after cranking some very loud music for the past hour. Both fets are screwed to it. I will bet that most of the heat was radiated from the nearby tubes. Pick a resistor between 0 and 4.7K that fits your board. Use a value closer to 4.7K if you have a small heat sink.
I cranked about an hours worth of classic rock through the board with the volume up full. Yes it was loud, real loud with 87db speakers. This weekend this amp will meet my 96db speakers with 15 inch woofers! The sound is very dynamic, snare drums are realistic, something that a SE amp just can't get right. All is not perfect, more gain is needed. I was using my $30 Walmart DVD player with a scope across both speakers. I never saw the amp clip once even with some ELP and Rick Wakeman. I saw 60 volts peak to peak across my speakers, but it was clean. Never heard any distortion either except for my 7 inch speakers trying to deal with Keith Emerson.
I think that 100WPC is probably the safe limit for continuous operation with these tubes and trasnformers. Music cranked up to where it clips on peaks is less demanding and should be no problem.
There are two power transformers in my set up. One has the red, blue, and green wires powering the PC board. It was barely warm after 1 hour of sine wave testing and 1 hour of loud music. The other transformer is using only the red leads and it was cold after all testing.
It remains to be seen how these tiny OPT's can deal with the 15 inch speakers and some techno music. If 2 of them can't cut it, I'll try 4!
The sockets in my board were from before PC board 12 pin sockets were commonly available, so I took chassis mount sockets and cut the pins in half length wise. Stan at ESRC had some PC board sockets so I got some last time I was there. Haven't tried them yet but they will go in the next build. Do not use the black plastic sockets from AES. They start smelling funny after a few minutes and became brittle after a month or two.
The tubes are 6HJ5's. 4 new tubes taken right out of their boxes. The idle current drifted around a bit for the first hour or so but is stable now. They had been sleeping for 35 years or so.....
The primary impedance can be anything from 6.6K to 3.3K. You will get about 75 WPC at clip with a 6.6K OPT and 135 WPC or so with a 3.3K. The distortion numbers were slightly better with the 6.6K but not much difference really. I haven't had a chance to listen carefully to both ratios. If you want maximum power, use the 3.3K OPT. I don't have any 3.3K OPTs at this moment so I am using some 6.6K transformers that are really too small for this amp. I can get 3.3K by hooking my 8 ohm load to the 16 ohm tap. Another possibility is a 4.2 or 5K OPT. I am guessing that 5K will get 85 to 95 WPC and 4.2K 110 to 120 WPC. The lower impedance runs the output tubes harder, but no issues were seen in a 5 minutes at 100WPC test with 3.3K.
I am going to try 2 OPT's per channel just to see what happens, but that is just me being too cheap to buy new transformers and too impatient to wait for them. Almost got a Hammond cheap on Ebay but got zonked in the last 10 seconds by a sniper.
I am currently using 6GU5's for the driver. They plug right in and give a bit more gain than the 6CB6. I will experiment some more when I have time to extract a bit more gain from this amp. I may try some different driver tubes, but I will stick with something that plugs in and costs $3 each or less.
Just my two cents, since I've been asked this privately a few times now, I am probably going with the CXPP100-MS-4.2K from Edcor for this amp (6HJ5 DCPP). My B+ will be lower than what George is currently running...probably a bit under 500V using the Antek transformer that I have.
This project has been bumped down by one or two other amp projects, so I've been resisting the temptation to fire it up again in light of this recent activity. I'm trying to finish what is on my bench, which has a another rather large chassis on it right now.
This project has been bumped down by one or two other amp projects, so I've been resisting the temptation to fire it up again in light of this recent activity. I'm trying to finish what is on my bench, which has a another rather large chassis on it right now.
Great stuff here! I've had my board for quite a while now, but thanks to two shoulder surgeries last year I haven't exactly been able to get to it. However, the shoulder is feeling much better so it won't be long now!
Just wondering how many folks have used this amp with Khorns? I have four of them and am curious if the consensus is the amps sound better in the original 18 watt configuration or some of the higher power configurations.
Just wondering how many folks have used this amp with Khorns? I have four of them and am curious if the consensus is the amps sound better in the original 18 watt configuration or some of the higher power configurations.
Last edited:
I have stuffed sweep tubes of all sizes into this amp and find that the sound characteristics don't change too much from 15 WPC to 250 WPC provided that tubes designed for TV horizontal sweep use are installed. I had the amp running with 6HE5's which are vertical sweep tubes. They need at least 250 volts on G2 but sound more like 6V6's. I would say figure out how many watts you need add plenty of margin and then pick the output tubes and transformers to get there. The board is rock solid at power levels up to 125 WPC.............
OK, I haven't been here for the last several hours because I just completed a major torture test on my latest creation, and tested the patience of my neighbors. I have been hesitant to really crank the red board into my Yamaha NS-10M's for fear of finding the woofer cone in the next zip code.
This afternoon I dragged the amp into the living room and hooked it up to the Silver Iris. These are 15 inch coaxial drivers able to take what I can feed them and are rated at 96db sensitivity. I have them mounted in a matched pair of 1941 vintage Zenith console radios, so there was some concern that the extreme power could dismantle the radio cabinets, but nothing bad happened.
I set the OPT's on the 6.6K ohm tap which gives about 75 WPC, connected the speakers and a decent DVD player. After warm up there was no hum at all detectable at the speakers, so I put on some techno and dimed the volume control. LOUD! Clean and solid. I played it this way for about half an hour and then switched to the 3300 ohm taps for about 125 WPC. The little 5 pound OPT's shouldn't like this. They could saturate on strong bass leading to blown output tubes, but no distortion was noted at full crank on some of the most bass heavy music I could find. The lower impedance produced a noticible but not big increase in volume and did not change the sound much.
I beat on it pretty hard for about two hours with no issues so having gained the confidence that the amp wasn't going to explode, I put on a Decyfer Down CD (LOUD metal with solid bass), opened the windows and went outside to mow my lawn. I could not hear the lawnmower! That particular CD is recorded loud enough so that the amp hits clipping occasionally. When the CD was done I touched each power transformer. The one that runs the board was warm. The one that makes only plate voltage was cold. The heat sink was cold. Only the tubes were hot. Ambient temp in the room was 69F which is colder than usual. I played some more music at a reduced power level for another hour of two.
So the amp was on non stop for 5 hours and was at maximum volume for at least 2 hours. After it got dark I shut off the lights and cranked some loud music for about 10 minutes. No glow was visible on the plates or screens.
So unless you are as crazy as I am, or even if you are, you aren't going to blow this amp up!!!!!!!!!
So what does it sound like. Well LOUD is obvious, but it is solid and the bass is clean even with my undersized OPT's. The sound is clean but not clinical or solid statey.
The Silver Iris are marketed toward the low powered SE crowd and I have been feeding them with a KT88 based Simple SE. They had big booming bass with the Simple SE. This is gone with the red board. There is less bass, but it is solid and well controlled. In fact if you want a rock concert in your living room, this is the combination to use. You'll need a biger living room!
I have a fresh unpopulated red board. I now have the combination for a reliable amp that I like, and will cost nearly nothing to build, so I guess I gotta git me one of these......
Hehe...cool George. I had the red board cranking hard for a few hours with the 6HJ5s at 475V. The heat sink got toasty, but not as bad as when it was eating sine waves. The tubes hardly broke a sweat. I was playing some Armin van Buuren for the bass, but the misses got cranky. 
If I ever do finish my queue of amps...I'm going to have an awful lot of them!
If I ever do finish my queue of amps...I'm going to have an awful lot of them!
Thanks George!
I guess I should have more properly inquired if any of the higher power versions changed the all important first watt. My Khorns are all modded with customer crossovers of my own design (points changed to 300hz and 4.5khz), running EV T-350 tweeters and EV HR9040 white whale midrange horns with EV DH1012 drivers. So, I guess they aren't really Khorns anymore, but there sure do sound smooth and, well, you get the idea. If there is any noise in the amplifier, it will be revealed. Bad recordings will drive you out of the room. Good recordings will have you streaming tears of joy.
I am currently powering them with 4 Bogen DB-130 monos that have been reconfigured into power amps and running 6FW5 output tubes with a 6CG7 front end. The 6FW5's run with -50vdc on the grid, 690 to 700vdc on the plates and a regulated 210vdc on the screens. Nary a single tube failure in over two years of daily operation. You can hear some amp hiss at about a foot or so away from the T-350's.
Hence, this is my curiosity with the red-board amp. If I can get the same or a little more power than the Bogen DB130's and also have a reduced noise floor, then that is a win-win situation!
The extra power would be cool for when I feel like cranking Bela Fleck and the Flecktones "Flight of the Cosmis Hippo", but I digress, Crazy? Me?
I guess I should have more properly inquired if any of the higher power versions changed the all important first watt. My Khorns are all modded with customer crossovers of my own design (points changed to 300hz and 4.5khz), running EV T-350 tweeters and EV HR9040 white whale midrange horns with EV DH1012 drivers. So, I guess they aren't really Khorns anymore, but there sure do sound smooth and, well, you get the idea. If there is any noise in the amplifier, it will be revealed. Bad recordings will drive you out of the room. Good recordings will have you streaming tears of joy.
I am currently powering them with 4 Bogen DB-130 monos that have been reconfigured into power amps and running 6FW5 output tubes with a 6CG7 front end. The 6FW5's run with -50vdc on the grid, 690 to 700vdc on the plates and a regulated 210vdc on the screens. Nary a single tube failure in over two years of daily operation. You can hear some amp hiss at about a foot or so away from the T-350's.
Hence, this is my curiosity with the red-board amp. If I can get the same or a little more power than the Bogen DB130's and also have a reduced noise floor, then that is a win-win situation!
The extra power would be cool for when I feel like cranking Bela Fleck and the Flecktones "Flight of the Cosmis Hippo", but I digress, Crazy? Me?
I only have 300 volts on the board so the fets don't have to eat as much. The 600 volt supply only goes to the OPT's. Nothing on the board needs 600 volts.
I have been experimenting with the idle current. At 600 volts it doesn't take too much current to put the tubes into the red zone. Todays experiments were done at 35 mA per tube. They were pretty dang hot!
I guess we now work for two different companies. I found one new glitch already. On Jan 4, I got a patent issued to MMI, I work for MSI. I bet I never see the issue money. Still working on LTE though.
"I had the amp running with 6HE5's which are vertical sweep tubes. They need at least 250 volts on G2 but sound more like 6V6's. "
So those do sound something like 6V6s then. The curves look more like 6L6s than 6V6s to me. (lower screen current near the knees, squarer) Wonder what the 6/10JA5 sound like. They list as a vertical amp tube, but have the lower screen voltage of a horiz. sweep. Curves look like H sweeps too.
Have you tried the 21HB5A Horiz. sweep tube? These list for a $1 and have a plate the same size as the 6HJ5. But a slightly smaller cathode. Only 18 watt rated.
So those do sound something like 6V6s then. The curves look more like 6L6s than 6V6s to me. (lower screen current near the knees, squarer) Wonder what the 6/10JA5 sound like. They list as a vertical amp tube, but have the lower screen voltage of a horiz. sweep. Curves look like H sweeps too.
Have you tried the 21HB5A Horiz. sweep tube? These list for a $1 and have a plate the same size as the 6HJ5. But a slightly smaller cathode. Only 18 watt rated.
Last edited:
There are two issues at hand here.
Hiss can come from a few places. The most likely is a noisy tube or resistor in the input stages of the amp. Another possibility is noise on one of the power supplies. The power supplies in the red board are mosfet regulated and filtered and quite clean. THere are dozens of different tubes that can be used in the input section of the red board, so a clean set should be not a problem. I just fired my amp up again and I hear no hum or hiss with my ear next to the speaker.
The second issue affects the sound quality of the first watt. As I mentioned in my last post as the B+ voltage goes up the safe current level goes down. The 6FW5 is a 17 watt tube. With 700 volts of B+ you will reach the disipation limit at 25mA. You need enough current to eliminate crossover distortion, but not enough to melt the tube.
I am running 600 volts and 24 watt tubes, which yields a maximum of 40 mA. 35 seems to sound good, but I may try some 35 watt tubes to see if there is anything to gain with higher current.
I do like your approach. I already have the AN-4TK400 slated for duty there. I'll just be burning more watts as heat than you.
I have some fun ideas for chassis in mind for the red board...D'oh. Yeah I bet there are going to be some snafus. Hope they sort it out. Funny that you work on LTE. So do I, now.
For the input section I picked up a few 6DK6 tubes from ESRC for $1 a piece which are compatible with the 6CB6. I have everything on the board but the sockets, fets, and schade feedback resistors(coming from mouser). I did accidentally lift a pad on the component side of the board when removing a capacitor due to me accidentally installing it backwards.