RoeTest tube testing system
- Thread starter pac1085
- Start date
Haven’t touched this project in a while but finally got around to building the calibration box over the last 2 days. The two small resistors on the 47k cement one were added since the 47k was a 10% part and way on the high side. I added two 3.3M resistors in parallel to drop it down a bit. May swap them out for a single 1.5M though.
Sure am glad I ordered everything when I did considering DHL stopped shipment into the USA for now. As far as I know Reichelt only uses DHL for US shipment, and they are the only one that seem to sell some of the parts used in this kit (like that male connector on the calibration box) - doesn’t seem to be available at all in the US.
Sure am glad I ordered everything when I did considering DHL stopped shipment into the USA for now. As far as I know Reichelt only uses DHL for US shipment, and they are the only one that seem to sell some of the parts used in this kit (like that male connector on the calibration box) - doesn’t seem to be available at all in the US.
Nice. You are so right on the shipping thing. Some parts can be sourced in the US, you may be able to find some substitutes in the US, which means you have to do some work to find suitable subs, but some parts aren't sourceable in the US. And it's mostly just easier to follow Helmut's recipe and get everything from Reichelt. Reichelt is great. Good prices, excellent service.
Now, though, getting things shipped from them has become much more difficult and expensive. I held off ordering the parts for the sound test accessory box. I finally submitted the order, but Reichelt/DHL aren't shipping out of the EU. You have to arrange for your own shipping. I'm going to give this a shot with my own shipping accounts and see how this goes. It's not just Reichelt and DHL. Other vendors from other markets are also stymied with the recent events. I've also already paid higher prices for parts from Digikey, Mouse, and Newark...
The sound test box isn't complex, but getting direct subs isn't so easy and I also wanted to order other bits, like DIN connectors and boxes to build a couple more socket adapter boxes and some other miscellaneous things to have as spares, and these parts are kind of effectively unobtainium.
Now, though, getting things shipped from them has become much more difficult and expensive. I held off ordering the parts for the sound test accessory box. I finally submitted the order, but Reichelt/DHL aren't shipping out of the EU. You have to arrange for your own shipping. I'm going to give this a shot with my own shipping accounts and see how this goes. It's not just Reichelt and DHL. Other vendors from other markets are also stymied with the recent events. I've also already paid higher prices for parts from Digikey, Mouse, and Newark...
The sound test box isn't complex, but getting direct subs isn't so easy and I also wanted to order other bits, like DIN connectors and boxes to build a couple more socket adapter boxes and some other miscellaneous things to have as spares, and these parts are kind of effectively unobtainium.
I assembled some of the socket boxes last night (I built them a while ago but never wired em up)
I should probably ground the two screws that attach the din connector but there’s no way to touch them when it’s installed, so I didn’t bother.
I only have 3 more of these din connectors left and won’t be able to order more any time soon, so I’ll have to build more boxes sparingly - or build a few sockets into one box assuming it doesn’t cause oscillation issues.
I wanted to keep the wires as short as possible since the ferrite beads to prevent oscillation are on the female side of the socket in the main unit.
Not the prettiest but they will work for now. In hindsight I’m thinking I should have used stranded wire, at least for the 7 and 9 pin since they move around a bit when installing/removing tubes. Eh, it’ll be fine. My tek 570 from the 50s has solid core wire to the pin sockets and hasn’t broken yet.
I think I might be able to get this thing fired up over the weekend!
I should probably ground the two screws that attach the din connector but there’s no way to touch them when it’s installed, so I didn’t bother.
I only have 3 more of these din connectors left and won’t be able to order more any time soon, so I’ll have to build more boxes sparingly - or build a few sockets into one box assuming it doesn’t cause oscillation issues.
I wanted to keep the wires as short as possible since the ferrite beads to prevent oscillation are on the female side of the socket in the main unit.
Not the prettiest but they will work for now. In hindsight I’m thinking I should have used stranded wire, at least for the 7 and 9 pin since they move around a bit when installing/removing tubes. Eh, it’ll be fine. My tek 570 from the 50s has solid core wire to the pin sockets and hasn’t broken yet.
I think I might be able to get this thing fired up over the weekend!
Got some more done today. Used fish paper to insulate the back side of the pcb.
Regrettably I broke a drill bit off while drilling one of the top holes to mount the heat sink to the front plate. Can’t get it out. Not the end of the world, as the heatsink is still held to the top plate with 3 other screws, but I’m gonna try to fix it anyways. That’s what I get for doing it with a hand drill. I knew better but I did it anyways.
I ordered this carbide drill bit set from KnKut, hopefully the 1/8” one does the trick. I also ordered an m4 bottoming tap to do a better job.
I’ve wanted a set of carbide bits for a long time, mostly from automotive work… so I don’t mind spending the $$
So, all that’s left now is wiring the test socket to the mainboard (I need to disassemble it again to do that), drilling and tapping a few more holes, including the larger rear fan…
Soldering the mosfets on and bolting them to the heatsink
And then pulling all of the boards out and bringing it up piece by piece per the manual.
Regrettably I broke a drill bit off while drilling one of the top holes to mount the heat sink to the front plate. Can’t get it out. Not the end of the world, as the heatsink is still held to the top plate with 3 other screws, but I’m gonna try to fix it anyways. That’s what I get for doing it with a hand drill. I knew better but I did it anyways.
I ordered this carbide drill bit set from KnKut, hopefully the 1/8” one does the trick. I also ordered an m4 bottoming tap to do a better job.
KnKut 5 Piece Carbide Tipped Hard Steel Drill Bit Set - Core Drill Bits - Amazon.com
KnKut 5 Piece Carbide Tipped Hard Steel Drill Bit Set - Core Drill Bits - Amazon.com
a.co
I’ve wanted a set of carbide bits for a long time, mostly from automotive work… so I don’t mind spending the $$
So, all that’s left now is wiring the test socket to the mainboard (I need to disassemble it again to do that), drilling and tapping a few more holes, including the larger rear fan…
Soldering the mosfets on and bolting them to the heatsink
And then pulling all of the boards out and bringing it up piece by piece per the manual.
Last edited:
The KnKut carbide tipped 1/8" bit made quick work of that broken off drill bit, cool! I didn't do a perfect job at removing it, there's still a piece broken off deep down below but I had plenty of threads to get the screw into so I just left it.
Well, I "finished" it last night, powered it up, added the boards etc.
I appear to have made two mistakes. The second one appears to have bit a bunch of other people as well, so I'm not counting it against me lol.
First mistake, I accidentally had the PCF8574N(i2c expander) and ULN2003A (darlington transistor array/relay driver) chips swapped with each other on relay card 1. After some swapping around, I determined that I cooked the PCF8574N chip. I need to order a replacement.
Second, for whatever reason I was shipped the SO8 SMD version of the LT1013 opamps used on the G1 and G3 cards. I believe these are used to measure voltages for the virtual meters inside the software. So, I was unable to calibrate those cards correctly or test any tubes. I didn't think much of the SMD version of the chip, just soldered it to an adapter board and installed it. Turns out the SO8 version has a totally different pinout than the DIL8 version - they call this out in the data sheet.
I could etch some adapter boards and move the pins around but I'm just going to order the correct LT1013CN8 chips.
I have a 3 day weekend (Monday off) so im coughing up the $46 shipping fee for UPS Overnight Saturday delivery from digikey. I'm pretty excited and want to play around with this thing all weekend!
Edit: parts ordered. I ordered a transformer to build the sound test box as well (he just has you use a 2x115 / 2x18v power transformer)
Well, I "finished" it last night, powered it up, added the boards etc.
I appear to have made two mistakes. The second one appears to have bit a bunch of other people as well, so I'm not counting it against me lol.
First mistake, I accidentally had the PCF8574N(i2c expander) and ULN2003A (darlington transistor array/relay driver) chips swapped with each other on relay card 1. After some swapping around, I determined that I cooked the PCF8574N chip. I need to order a replacement.
Second, for whatever reason I was shipped the SO8 SMD version of the LT1013 opamps used on the G1 and G3 cards. I believe these are used to measure voltages for the virtual meters inside the software. So, I was unable to calibrate those cards correctly or test any tubes. I didn't think much of the SMD version of the chip, just soldered it to an adapter board and installed it. Turns out the SO8 version has a totally different pinout than the DIL8 version - they call this out in the data sheet.
I could etch some adapter boards and move the pins around but I'm just going to order the correct LT1013CN8 chips.
I have a 3 day weekend (Monday off) so im coughing up the $46 shipping fee for UPS Overnight Saturday delivery from digikey. I'm pretty excited and want to play around with this thing all weekend!
Edit: parts ordered. I ordered a transformer to build the sound test box as well (he just has you use a 2x115 / 2x18v power transformer)
Last edited:
I'm always so impressed with the build pictures of RoeTest. Thank you for posting all the updates and information. If you get a chance, I’d love to see some of the test results as well. I wonder how test results between the RoeTest and the Amplitrex AT1000 on the same tube compare. I know RoeTest is far more comprehensive. I have been messing with making my own software for the AT1000, going slow dont want to break anything. Thanks again!
Last edited:
I build the sound test transformer box today. It allows you to connect to an amplifier and tap on the tube to listen for Microphonics and noise. You also get to control all of the parameters of the test so you can simulate how it might behave in an actual amplifier. Works great!! It will take a bit of experience to get the hang of though, since it seems all tubes have at least some noise (though I did find some very quiet ones compared to others in some tests)
I also built an adapter for the UX4, UX5, UX6 tube bases. I had an 807 and an 83 I wanted to test. In a perfect world you'd build one adapter box per tube type, so you can keep the wires short to avoid oscillation. He already has large ferrite beads on the receiving side of the socket box. Unfortunately its impossible to order any more of the DIN 41622 12 pin male connectors as they are only available in Europe and nobody's shipping to the states at the moment. I bought 3 spares, but I will have to ration them carefully.
I may consolidate my 7-pin and 9-pin sockets into one box, and add a nuvistor socket to it as well. I have the Hickok SA-3 nuvistor adapter, it works well but you have to go into the Roetest database and duplicate/edit each tube type you want to use it with due to the pin assignments being different.
Automated magic eye test is neat too.
I also built an adapter for the UX4, UX5, UX6 tube bases. I had an 807 and an 83 I wanted to test. In a perfect world you'd build one adapter box per tube type, so you can keep the wires short to avoid oscillation. He already has large ferrite beads on the receiving side of the socket box. Unfortunately its impossible to order any more of the DIN 41622 12 pin male connectors as they are only available in Europe and nobody's shipping to the states at the moment. I bought 3 spares, but I will have to ration them carefully.
I may consolidate my 7-pin and 9-pin sockets into one box, and add a nuvistor socket to it as well. I have the Hickok SA-3 nuvistor adapter, it works well but you have to go into the Roetest database and duplicate/edit each tube type you want to use it with due to the pin assignments being different.
Automated magic eye test is neat too.
Last edited:
So far one of the most useful functions of the RoeTest has been its grid current measurements to identify gassy tubes.
I was using my Tek 570 last week and noticed the high voltage section was losing regulation causing the CRT picture to bloom badly. At first I was afraid it was a bad HV transformer. I had replaced the 6AQ5 tube in the CRT circuit a few months ago, and selected a good one using my Dan Nelson calibrated TV-7D/U. It measured no gas on the TV-7. However the RoeTest seems to be much better at measuring grid current, and it weeded out this tube as bad. I replaced the 6AQ5 with one that had an acceptable grid current reading and the 570 has been rock solid since.
I bought some 5933 tubes on eBay, I thought one of them looked suspicious because the Sylvania gold brand logo turned from gold to gray (where the other tube I had in the pair was still gold) - indicating it probably got REALLY hot. That tube ended up measuring a ridiculous amount of grid current on the RoeTest, but still nothing on the TV-7. Hmm...
I was using my Tek 570 last week and noticed the high voltage section was losing regulation causing the CRT picture to bloom badly. At first I was afraid it was a bad HV transformer. I had replaced the 6AQ5 tube in the CRT circuit a few months ago, and selected a good one using my Dan Nelson calibrated TV-7D/U. It measured no gas on the TV-7. However the RoeTest seems to be much better at measuring grid current, and it weeded out this tube as bad. I replaced the 6AQ5 with one that had an acceptable grid current reading and the 570 has been rock solid since.
I bought some 5933 tubes on eBay, I thought one of them looked suspicious because the Sylvania gold brand logo turned from gold to gray (where the other tube I had in the pair was still gold) - indicating it probably got REALLY hot. That tube ended up measuring a ridiculous amount of grid current on the RoeTest, but still nothing on the TV-7. Hmm...
arts
Addicted Member
Wow, good to hear it wasn't the HV transformer. I don't think they're a common model. I have a Tek 525 that I'm keeping around for spares (same CRT as my 570, just in case) but the HV transformer/circuit isn't the same.
I have a TV-7C/U that I don't really use, but I thought most of the Hickok designs were pretty good for gas tests. How much grid current was the tube showing?
I have a TV-7C/U that I don't really use, but I thought most of the Hickok designs were pretty good for gas tests. How much grid current was the tube showing?
1.377uA where a normal good tube is something like 0.0xuA
the 5933 tube I tested today measured 14uA!!! But not even 1 minor division of deflection on the TV-7. I wonder if it has to to with it being tested at a different operating point there, where as the RoeTest is testing it at the actual datasheet operating point. The gas test does work on the TV-7, I have a bunch of crappy 6DJ8's that it gets plenty of deflection on ...
targeteye
Super Member
Nice Write up on this pac1085 Thanks.
I bought the foundational kit earlier this year but i've been to busy to take it any further. Now i'm now working up the Reichelt order. Quite a bit of the parts are unavailable some no longer available. Did you encounter this? I'm pretty sure it can all be substituted but i and curious if you ran into the same issue.
Steve
I bought the foundational kit earlier this year but i've been to busy to take it any further. Now i'm now working up the Reichelt order. Quite a bit of the parts are unavailable some no longer available. Did you encounter this? I'm pretty sure it can all be substituted but i and curious if you ran into the same issue.
Steve
I tried to order parts from Reichelt earlier this year and they had stopped shipping to the US. Between the electronics components suppliers in the US - Mouser, Digikey, Newark, and some others, even EBay - I was able to source some things, like the various DIN connectors the RoeTest uses. There are other European suppliers, too. For some things I wasn't able to get exact substitutes. And the costs are all mostly much higher now.
Last edited:
targeteye
Super Member
i exported and imported.
I guess its gonna be a bit of a challenge to round up all those parts
Steve