Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Adapter PCBs for 148-003x-00 relay replacement
|
Hello all,
after quite a few years of absence from this group and from my trusty 7000-series equipment, I am back. The reason is quite obvious. I have a special measurement job which needs my 7A13. As expected, after almost 10 years of sitting around my 7A13 has quite a few issues, but this is a story on its own. As the subject already tells all the relays within the 7A13 have gone bad. Some seemed to work fine, but on closer inspection every relay has its own problems. Mostly contact resistance of the relays is highly unstable. Sometime you can get the relay down to 100mOhm, sometimes the same contact has up to 10 Ohm. Of course, I tried all the recommended reviving techniques like repeated switching at high frequency under load, etc. but nothing really helped. I remember that I replaced 2 of the relays already 10 years ago by a self-made free-floating solution to fit a new relay into the old socket. This was quite tedious and I don't want to repeat the same thing again. Leave alone the fact that this free-floating solution might not be capable of providing the right HF and Isolation performance. By searching for a better solution in this group I saw the solution of Tom Ford back in 2019. (/g/TekScopes/album?id=91736) Unfortunately, he didn't provide any design data and he seems to be offline since then. Therefore, I went to action and designed my own little adapter PCB. And I will also try to document my work as good as possible, to give anyone else enough information to do this on his own. Related pictures can be found here and maybe on my GibHub. /g/TekScopes/album?id=299851 My first Idea was to replace the internals of the original TEK relays with a modern relay with the original TEK Housing still in place. Therefore, the PCB had to be so small to fit inside the original housing and connect to the relay and the header pins. In addition to the relay there also should be a small bridge rectifier to get rid of the polarity issue most modern relays have (most modern relays are polarized) and optional a little SMD LED to see if the relay is activated. As the space within the relay is really tiny, I came up with the idea of using a 2layer Flex PCB which would be bent and hold the relay, a bridge rectifier and an LED + resistor. This concept (concept 1) looked really good but I had quite a few problems: 1. Crosstalk between channels and isolation between open contacts was not that good because of relatively long tracks in the flat flex running in parallel. 2. Flux residue between the flat flex and the header caused leakage which cause "enormous" leakage currents. Not the required +/-0.2nA more the +/-15nA or even more. 3. Due to the tight space within the relay the bending radius of the flat flex was quite tight and only in the range of 1-1.5mm. This caused on some trials problems with broken tracks. All in all, the ideas sounded great but the realization was not the great at all. For my next concept (concept2) I reverted back to the idea of Tom Ford and created my own little adapter PCB for both needed relay types used in the 7A13. At the moment this is an unproven design as the relays are still on back order and I'm waiting for the distributor to dispatch my order. For the relay I did a longer market study to find the best fitting relay. In the end I decided to use the AXICOM IM06TS which is really tiny and still provides relatively good switching and HF performance. The footprint of the relay is a little issue as the row spacing between the pins is only 5.08mm. This makes the pads slightly interfere with the through hole pads of the socket header. To still accommodate the socket header and the relay on the same PCB without interfering I decided that the pins of the relay need to be bent outwards a little (~.7mm per side). This is not a really nice solution but I assume it will work fine. If the socket header would have been available in a SMT version the use of a SMT relay would have made everything a lot easier. I presume Tom Ford used the BAS3007 as bridge rectifier. Unfortunately, this part was not available at my favorite distributor. Thus, I selected the BAS40DW-04 which is not exactly a bridge rectifier, but it can be connected as one while still being really tiny. As I already wrote I also added a small surface mount LED together with a resistor to the PCB because I experienced on my unit some issues with the Cam-Switches which could have been easily debugged if there had been a little indicator in place. All in all, the BOM for a 148-0034-00 relay replacement looks like this, for the 148-0035-00 it is basically the same, you only need to remove some pins from the pin header rows. 1x Relay (IM06TS) 0.1x Pin Header (W+P 952-09-080-2 cut into 10 pieces) 1x SMD Resistor 0603 1.8kOhm 1x SMD LED 0603 yellow The resulting PCB is measuring 8 x 14.5mm and is still smaller than the original relay (8.5 x 16mm). My favorite Chinese PCB manufacturer charges me the same no matter if the PCB is 8x14.5mm or 100x100mm. Therefore, I decided to make my own custom panel which utilizes the 100x100mm as good as possible. Now I have 5 panels of 66PCBs each lying next to me waiting for assembly. Each PCB holds 36 148-0034-00 compatible PCBs and 30 148-0035-00 compatible PCBs. As already stated above I was not able to do any tests, but I will definitely do tests as soon as I have some relay. As soon as my concept of the PCB is tested and proven good, I will give you all a follow up on the status and will also publish the design data in case you want to replicate or improve my design. I will not need all 330 PCBs, maybe 15 to 20. Therefore, I am willing to sell these to anyone interested in getting some relay replacements. As I also ordered a solder paste stencil I will wait how big the interesst in these PCB is before i start populating a whole panel. I think it would be the easiest if I sell these PCBs with the SMT parts already soldered on. Best regards, Wolfgang |
|
On Fri, Jan 3, 2025 at 08:02 AM, <w.m.d.koch@...> wrote:
I will not need all 330 PCBs, maybe 15 to 20. Therefore, I am willing to sellI'm certainly interested, at least in the PCB. I've actually been having this same battle with a 7A13 this last week, and was getting read to ask the list if anyone had published a PCB adapter since Tom Ford's isn't available. I think I have at least two that are causing different problems, but hate to cut into them for cleaning without a backup plan in place. Some of the relay alternatives noted in the archives are also obsolete, and I don't think any of them are rated to the original 7A13 spec of 500 V (though I may be interpreting that wrongly). thanks, Adam |
|
I'd like to have some of those boards. I have two 7A13s and would like to replace all the relays. I also have a 7A12 which uses 8 of them.
toggle quoted message
Show quoted text
In addition I need two of the "polarized" relays (148-0063-00). I cobbled up something out of two small relays that worked but getting two small relays in the small compartment is challenging. Please let me know. Thanks, Barry - N4BUQ ----- Original Message -----
From: "w.m.d.koch via groups.io" <w.m.d.koch@...> Hello all, |
|
After seeing some interest, I decided to populate at least 1 panel with all SMT components. This might also be the way selling them to you. This enables me to paste and reflow solder a whole panel which makes soldering of the SOT363 rectifier component a lot easier than soldering a single tiny PCB by hand.
As said in my initial post, I'm still waiting for the relays to arrive. Then I'd also like to do some measurements to see how good the solution performs in comparison to the original relay. Therefore, I think it takes at least another 2-3 weeks until I can give you some feedback. At work I have an 8GHz VNA (which is hopefully not blocked by someone else), this might be helpful when exploring HF-properties of the whole contraption. Regarding the 148-0063-00 relay. I had to google the pinout of this thing, but I think this should also be doable with a little solder bridge between adjacent pins. Best Regards, Wolfgang |
|
Hi Wolfgang,
toggle quoted message
Show quoted text
Regarding the 148-0063-00, I don't know what type of relay you're thinking of using but that relay is designed to operate such that when the polarity is one way, one set of contacts are closed and when the polarity is reversed, the other set of contacts are closed. I replicated that action with two relays but if you know of a dual-polarity relay like that, I'd definitely be interested. Thanks, Barry - N4BUQ After seeing some interest, I decided to populate at least 1 panel with all SMT |
|
Hi Barry,
sorry for the confusion regarding the 148-0063-00. I didn't know the part number and just googled the part. From the pictures I found I thought it was just as simple DPDT relay (like 148-0034-00) with an additional internal connection. As it turns out, this is not the case. It might be possible to build an PCB adapter for this one as well. But as it is only used in the 7A12 and nowhere else it might not be feasible to go through this effort. Bye, Wolfgang |
|
Hi Wolfgang,
toggle quoted message
Show quoted text
I understand. I plan to pursue those relays on my own but just wanted to ask just in case. I'm still interested in getting a few of your boards as I have two 7A13s and I think the 7A12 also uses some of those types as well. Thanks again, Barry - N4BUQ Hi Barry, |
|
Hello everyone,
its been a while since my last post. There were a lot of things which kept me from posting an update. Nevertheless, I finally managed to do some measurements of my relay replacements. The VNA at work was blocked all the time. Therefore, I had to find a different way to measure the relays. In the Album there are now the measurement result of almost all possible combinations you can have in a DPDT relay. /g/TekScopes/album?id=299851 All in all the replacements perform quite a bit worse than the originals. This is not surprising as the original relays are close to perfect in terms of HF coupling (tiny cross sections with relatively large gaps between adjacent parts). The PCB in contrast provides 'long' traces in parallel. As the VNA was not available I had to measure the relays with a automated setup composed of my 60MHz function generator as signal source and my digital oscilloscope as measurement device. Due to the noise floor of the oscilloscope and the low output level of the funtion generator dynamics of the measurement was not crazy high. To get an impression how much coupling would be possible at the given distances between pins, I always measured one complete trace without any DUT at all (the traces called 'Fixture'). Nevertheless, the measurements always show that the replacement performs not as good as the original relay, but I think this is the price you will have to pay to get a relaibly switching relay. Regarding leckage current I will try to perform a few measurements to see if the required 0.2nA is achivable or not. I hope the next update will not be delayed as much as this one. Right now I have 4 unpopulated panels left over (each with 36x 148-0034-00 and 30x 148-0035-00 relays). Plus some additional ones of the first panel which were already populated with SMT components. If you are interested, send me a message off list with your needs and we can discuss how I can help you. Best regards, Wolfgang |
Jean-Paul
to navigate to use esc to dismiss