Geez Frankie, I have to give it to you for paitence and tenacity, qualities very few possess these days. I looked up that jFET, it is something like 2N4391, 2N4392 or 2N4393….although not specifically listed in the HP part translation lists, 1855-0414 is a 2N4393 and the next listed is 1855-0420, which is a 2N4391…then those in berween are most likley to be SELECTED versions….i.e. somone at HP has sat down in a little room with a curve tracer and gone through a whole heap of 2N439x FETs and simply selected ones that wotk the sbsolute best in this circuit. If you look in the circuit, you can see it is switching the signal level at the output of the amplifier composed of Q17A/B, Q11, Q12, Q13, Q18 and transistors inside U1 (D and E). the output comes out via a 4K64, (R11), and then either has 511 Ohms (R20) to ground if Q19 is on or no R20 if Q19 is off. What they would have selected for is rds-on, because it adds in series with R20. If you can find a FET with a lower rds-on, then you can simply trim-up by adding resistors in series with R20 or just increasing the value of R20….its just to get the signal level spot-on a certain value when Q19 is turned on. I have tons of 2N4093’s here if you want any of those, the main difference being that the 2N4393 can handle 50mA forward bias gate current, whereas the 2N4093 is only about 10mA. The main difference between the 391, 392 and 393 is what I thought, rds-on, of 30, 40 and 50 ohms respectively, same for the 091, 092 and 093. That 4% beryllium 96% copper is amazing stuff, but, yes it fatigues quite easily…i.e. breaks when bent back and fourth. The other property the nickel did not have was lack of elastic limit….my father is a mechanical engineer and talks about this stuff all the time…so you can bend a material…and if it has a high elastic limit and you don’t bend it to that limit, it will spring back to its original shape….this is why they put the 4% Be in the copper, to increase its elastic limit…just thought you’d like to know the term for it! Can’t believe your rotten lick with damaged front pannel controls…my Tek 7L5 had a badly smashed up beryllium copper fingered PCB based span per division switch…but I managed to cobble it sll back together. Watched you go through the spectrometer and was impressed by how you used zero blaze to zero it and calibrate it. Thes big companies that make such stuff…waaayyy underestimate the ingenuity of some individuals such as youreslf. Watched the autoclaves too, so say hello to BrAD and Clare for me…..scrape some scale out of their boilers for this bloke in Tassie!
The Globe Collector Returns! I hope you are doing well and that everything is good on your side of the globe! I was starting to miss your detailed analysis of my videos. I always learn so much from them. Thanks for that info regarding Q19 on the Vertical Driver board! Per the service manual, Q19 1855-0417 is on when Linear or 10dB/div is selected, and off when dB/div is selected. I was able to find the electrical characteristics on CuriousMarc’s website! Link here (drive.google.com/drive/folders/1e4pSjjSeVk8JM6O1EnnuL9jYS6FAK8hN). And you’re right: the on resistance for 1855-0417 is low at only 10ohms. It seems like 2N4391 would be a good alternative. Right now I have a J113 in it’s place and the unit still seems to work, but its Ron is 100ohms. I think I might do a J111 instead (I believe I have some of those). I also appreciate your offer for sending over some JFETs! Finding the correct material for those contacts was the most difficult part of this repair! And thank you for providing the correct material characteristic terms. Elastic limit is definitely better than ‘springy’ lol. Still curious to know what the originals are made of though because they seem a lot more durable than the CuBe fingerstock I used. Any thoughts on that? When I was looking at SAs to buy, I saw a 7L5 plug-in for sale in non-working condition. I almost bought it, but I would have had to grab a 7000 series mainframe as well. Maybe someday! Thanks for watching the spectrophotometer and autoclave vids as well! It felt good to solve that spectrophotometer issue. I enjoy sticking it to these manufacturers lol. And I’ll be sure to get those boilers properly flushed!
Oh, yes…the 7L5…I slmost forgot. I have two of them, one a froend bought to the door one evening snd I got right into it. It is s fiendishly complex circuit. Each individual circuit block is dort of easy enough to grasp, butbthevway thay all “talk to each other” is sheer digito-analog wizardry! I ended up buyong a second non-functional unit off fleabay just so I could swap functionsl blocks aroind between the two. I think the problem is in one or both of two sample and hold circuits that hold the centre freqiency horizontal position voltage. They use fancy Analog Devices opnamps with a swanky, low leakage capacitor in the feedback loop…this is supposed to hold about 99% of its charge for at least ten seconds according to the manual. In the unit my friend bought around…after fixing smashed multi position switches, bent pot shafts, shorted Tantalum caps, open circuit rail dropper resistors..I found that the outpus of both S and H’s were slammed to the negative rail. The charge is added to or removed from the capacitors via two ultra low leakage diodes inna TO-72 can on the input of the S and H circuits…these diodes sre still svailable and thay sre the ONLY diodes with this low leakage as far as I can tell…a dizzyingly tiny 1pA! About three orders of magnitude less than most other diodes! Currently I am procrastinating getting back onto them, the second fleabay unit is just sitting in there on top of the other one basically untouched. Still a bit pooped after the container move snd its the bloody middle of winter here in Tas, day temps around 50 pf your antiquated Farenheight and night temps around 40…. It is cycling between grey, drizzly and sopping wet to sunny and windy every frw days. Sunrise is 8.45 in the morning snd sunset is 4.30 in the afternoon. The day is gone before it starts! Glad the winter solstice has passed and the sun is slowly starting its southward move again. Mut be noce there in Philly though at the summer solstice. Go for a walk to the Stocks Bakery and get some of that cake and eat it on video…lot easier than fixing electronic gear! I went an bought some “Madeira Cake” at Coles…it weighed 450g, and since an imperial pound is 454g, I see why it is called “pound cake” and I suspect this is the nearest equivalent I can get here. Gave one to my father all covered in whipped cream as an 88th birthday cake, he ate about 65% of it in 15 minutes! Get back to you when I take the 7L5 “bull by the horns” again!
@@thanhhuynh272 I'll get to that bakery for you eventually! I think that will make for an interesting video... "what's he doing eating cake? I thought this was about repairing electronics" 😂😂
Hi Frankie I have just got 2 of these plugins but minus the big knobs on the left hand side, I will have to make some out of aluminum on the lathe. Would appreciate a blown up picture of these as they are so complex.
Hi Frankie,it,s the attenuation/ref level/variable and the bandwidth /resolution they seem to have concentric shafts all very complex. Also yours has a mechanical freq readout mine have electronic display (dont think its a counter just a DVM display). Thanks
@@mechmania8450 gotcha. Is yours a 8557A? I know some of the 8557s had digital readouts. So you need knobs for the Optimum Level, Coarse Reference Level, and Fine Reference Level. As well as knobs for the Frequency Span/Div + Bandwidth Res? Thats 5 separate knobs. Am I understanding you correctly? If not why don't you shoot me an email (will need a way to send you photos anyway). My email is in the bio of my channel
You might get some additional help by looking at the service manual. It has an exploded diagram of the front-end. Because yes both those switches are concentric. But one is tri-stacked (amplitude) and the other double-stacked. You correct it is a pretty complex set of switches.
@@fmashockie Hi Frankie Cannot find you email. I have a scale for the attenuation/ref level and I am making a knob in aluminum. However I am not share how the bandwith/sweep width works there is a hole on the panel. Had me confused for a bit with the atten knob which has a channel front to back however this is not needed but it must weaken the knob. But I still wonder how the bw/sweep works. I have repaired helipots in the past ,bad connection at end of wire wound track.. Use some agressive soldering flux.
Great video Frankie. Enjoyed the trouble shooting process. Keep up the good work.
Thanks Joel! This one really tested my patience but I think the worst is behind me 😂😂 Always appreciate your feedback!
Geez Frankie, I have to give it to you for paitence and tenacity, qualities very few possess these days.
I looked up that jFET, it is something like 2N4391, 2N4392 or 2N4393….although not specifically listed in the HP part translation lists, 1855-0414 is a 2N4393 and the next listed is 1855-0420, which is a 2N4391…then those in berween are most likley to be SELECTED versions….i.e. somone at HP has sat down in a little room with a curve tracer and gone through a whole heap of 2N439x FETs and simply selected ones that wotk the sbsolute best in this circuit.
If you look in the circuit, you can see it is switching the signal level at the output of the amplifier composed of Q17A/B, Q11, Q12, Q13, Q18 and transistors inside U1 (D and E). the output comes out via a 4K64, (R11), and then either has 511 Ohms (R20) to ground if Q19 is on or no R20 if Q19 is off.
What they would have selected for is rds-on, because it adds in series with R20. If you can find a FET with a lower rds-on, then you can simply trim-up by adding resistors in series with R20 or just increasing the value of R20….its just to get the signal level spot-on a certain value when Q19 is turned on.
I have tons of 2N4093’s here if you want any of those, the main difference being that the 2N4393 can handle 50mA forward bias gate current, whereas the 2N4093 is only about 10mA. The main difference between the 391, 392 and 393 is what I thought, rds-on, of 30, 40 and 50 ohms respectively, same for the 091, 092 and 093.
That 4% beryllium 96% copper is amazing stuff, but, yes it fatigues quite easily…i.e. breaks when bent back and fourth.
The other property the nickel did not have was lack of elastic limit….my father is a mechanical engineer and talks about this stuff all the time…so you can bend a material…and if it has a high elastic limit and you don’t bend it to that limit, it will spring back to its original shape….this is why they put the 4% Be in the copper, to increase its elastic limit…just thought you’d like to know the term for it!
Can’t believe your rotten lick with damaged front pannel controls…my Tek 7L5 had a badly smashed up beryllium copper fingered PCB based span per division switch…but I managed to cobble it sll back together.
Watched you go through the spectrometer and was impressed by how you used zero blaze to zero it and calibrate it. Thes big companies that make such stuff…waaayyy underestimate the ingenuity of some individuals such as youreslf.
Watched the autoclaves too, so say hello to BrAD and Clare for me…..scrape some scale out of their boilers for this bloke in Tassie!
The Globe Collector Returns! I hope you are doing well and that everything is good on your side of the globe! I was starting to miss your detailed analysis of my videos. I always learn so much from them.
Thanks for that info regarding Q19 on the Vertical Driver board! Per the service manual, Q19 1855-0417 is on when Linear or 10dB/div is selected, and off when dB/div is selected. I was able to find the electrical characteristics on CuriousMarc’s website! Link here (drive.google.com/drive/folders/1e4pSjjSeVk8JM6O1EnnuL9jYS6FAK8hN). And you’re right: the on resistance for 1855-0417 is low at only 10ohms. It seems like 2N4391 would be a good alternative. Right now I have a J113 in it’s place and the unit still seems to work, but its Ron is 100ohms. I think I might do a J111 instead (I believe I have some of those). I also appreciate your offer for sending over some JFETs!
Finding the correct material for those contacts was the most difficult part of this repair! And thank you for providing the correct material characteristic terms. Elastic limit is definitely better than ‘springy’ lol. Still curious to know what the originals are made of though because they seem a lot more durable than the CuBe fingerstock I used. Any thoughts on that?
When I was looking at SAs to buy, I saw a 7L5 plug-in for sale in non-working condition. I almost bought it, but I would have had to grab a 7000 series mainframe as well. Maybe someday!
Thanks for watching the spectrophotometer and autoclave vids as well! It felt good to solve that spectrophotometer issue. I enjoy sticking it to these manufacturers lol. And I’ll be sure to get those boilers properly flushed!
Oh, yes…the 7L5…I slmost forgot. I have two of them, one a froend bought to the door one evening snd I got right into it. It is s fiendishly complex circuit.
Each individual circuit block is dort of easy enough to grasp, butbthevway thay all “talk to each other” is sheer digito-analog wizardry!
I ended up buyong a second non-functional unit off fleabay just so I could swap functionsl blocks aroind between the two.
I think the problem is in one or both of two sample and hold circuits that hold the centre freqiency horizontal position voltage.
They use fancy Analog Devices opnamps with a swanky, low leakage capacitor in the feedback loop…this is supposed to hold about 99% of its charge for at least ten seconds according to the manual. In the unit my friend bought around…after fixing smashed multi position switches, bent pot shafts, shorted Tantalum caps, open circuit rail dropper resistors..I found that the outpus of both S and H’s were slammed to the negative rail.
The charge is added to or removed from the capacitors via two ultra low leakage diodes inna TO-72 can on the input of the S and H circuits…these diodes sre still svailable and thay sre the ONLY diodes with this low leakage as far as I can tell…a dizzyingly tiny 1pA! About three orders of magnitude less than most other diodes!
Currently I am procrastinating getting back onto them, the second fleabay unit is just sitting in there on top of the other one basically untouched. Still a bit pooped after the container move snd its the bloody middle of winter here in Tas, day temps around 50 pf your antiquated Farenheight and night temps around 40…. It is cycling between grey, drizzly and sopping wet to sunny and windy every frw days. Sunrise is 8.45 in the morning snd sunset is 4.30 in the afternoon. The day is gone before it starts!
Glad the winter solstice has passed and the sun is slowly starting its southward move again. Mut be noce there in Philly though at the summer solstice. Go for a walk to the Stocks Bakery and get some of that cake and eat it on video…lot easier than fixing electronic gear!
I went an bought some “Madeira Cake” at Coles…it weighed 450g, and since an imperial pound is 454g, I see why it is called “pound cake” and I suspect this is the nearest equivalent I can get here. Gave one to my father all covered in whipped cream as an 88th birthday cake, he ate about 65% of it in 15 minutes!
Get back to you when I take the 7L5 “bull by the horns” again!
@@thanhhuynh272 I'll get to that bakery for you eventually! I think that will make for an interesting video... "what's he doing eating cake? I thought this was about repairing electronics" 😂😂
Hi Frankie I
have just got 2 of these plugins but minus the big knobs on the left hand side, I will have to make some out of aluminum on the lathe. Would appreciate a blown up picture of these as they are so complex.
Hey glad to help you out but which part are you referring to exactly?
Hi Frankie,it,s the attenuation/ref level/variable and the bandwidth /resolution they seem to have concentric shafts all very complex. Also yours has a mechanical freq readout mine have electronic display (dont think its a counter just a DVM display). Thanks
@@mechmania8450 gotcha. Is yours a 8557A? I know some of the 8557s had digital readouts. So you need knobs for the Optimum Level, Coarse Reference Level, and Fine Reference Level. As well as knobs for the Frequency Span/Div + Bandwidth Res? Thats 5 separate knobs. Am I understanding you correctly? If not why don't you shoot me an email (will need a way to send you photos anyway). My email is in the bio of my channel
You might get some additional help by looking at the service manual. It has an exploded diagram of the front-end. Because yes both those switches are concentric. But one is tri-stacked (amplitude) and the other double-stacked. You correct it is a pretty complex set of switches.
@@fmashockie Hi Frankie Cannot find you email. I have a scale for the attenuation/ref level and I am making a knob in aluminum. However I am not share how the bandwith/sweep width works there is a hole on the panel. Had me confused for a bit with the atten knob which has a channel front to back however this is not needed but it must weaken the knob. But I still wonder how the bw/sweep works. I have repaired helipots in the past ,bad connection at end of wire wound track.. Use some agressive soldering flux.