Keysight 3000T scope 1 GHz(or more) hardware upgrade mod

[Keysight DanielBogdanoff]

The 500 MHz mod for the T series should be roughly the same as the A series however I don't think anyone has ever actually done it to verify 100% the exact components that need to be changed.

I haven't seen it done either and agree that it would be roughly the same. There's not a lot of difference in signal paths between the A/T/G versions.

[TheSteve]

The 500 MHz mod for the T series should be roughly the same as the A series however I don't think anyone has ever actually done it to verify 100% the exact components that need to be changed.

I haven't seen it done either and agree that it would be roughly the same. There's not a lot of difference in signal paths between the A/T/G versions.

The 1.5 GHz mod is more fun, wouldn't you agree? 

[pquadrat]

Agreed. I'm impressed that people are willing to hardware mod such an expensive high end scope.

After Steve lead the way, for others should be child's play (except me, of course... ).

I did the 500MHz mod on a 3104T. Working perfectly. Before calibration, the bandwith was about 750Mhz, which gets "corrected" in calibration.

I did not even know that it was for the A model. Everything seems to be identical.

[maxwell3e10]

The 500 MHz mod for the T series should be roughly the same as the A series however I don't think anyone has ever actually done it to verify 100% the exact components that need to be changed.

I haven't seen it done either and agree that it would be roughly the same. There's not a lot of difference in signal paths between the A/T/G versions.

The 1.5 GHz mod is more fun, wouldn't you agree? 

To summarize the basic difference if  I understand it correctly: upgrading up to 500 MHz requires changing a few L, C and R per channel. The most difficult part is unsoldering the RF shields. To upgrade to 1 GHz requires changing relays and other multi-terminal components which are trickier to desolder.

[TheSteve]

Pretty much, the 1 GHz mod is a much more extensive upgrade - and it costs a lot more because of the Teledyne attenuator relays.

[maxwell3e10]

Pretty much, the 1 GHz mod is a much more extensive upgrade - and it costs a lot more because of the Teledyne attenuator relays.

It's good to know the price difference between 500M and 1GHz versions is not just pure mark-up 

[EE-digger]

Pretty much, the 1 GHz mod is a much more extensive upgrade - and it costs a lot more because of the Teledyne attenuator relays.

It's good to know the price difference between 500M and 1GHz versions is not just pure mark-up 

Not pure ... let's just leave it at that.  The R&D for a company the size of Keysight to develop a 1GHz scope is probably significant.  It was done a long time ago though, just like the Megazoom IV.

I had all parts on hand for the 1GHz mod and decided to try the 500MHz first, which was simple.  I've gotten used to 0402 and 0201 parts though so the component swaps were for BIG components.  I thought I had marked photos of the 3000T to follow.  Matter of fact, it was the 3000T because it was a 3014T but had footprints for the Teledyne relays.  3000A did not.

There are around 132 or so components for the 1GHz mod.  I would have done it but finally, a pristine 3104T appeared and with option bundle to boot.  Someone on here grabbed the parts just to get the 4 relays in the set.

[EE-digger]

Quick 2025 update.  Should have invested in relays instead of the market.  The relays were $60 a few years ago and now $150 from the same source.

[tv84]

Quick 2025 update.  Should have invested in relays instead of the market.  The relays were $60 a few years ago and now $150 from the same source.

Make Relays Great Again!

[TheSteve]

Painful. I seem to recall they were always expensive, and hard to get in small quantities. I couldn't get them in North America.

[EEVblog]

I have forgotten about this.
Does the 3000T PCB have a dedicated path layout for the 1GHz bandwidth input and the (132?!) components just left out?
And the 500MHz version uses another path that gets left unpopulated if it's the 1GHz model?

[analogRF]

Does upgrading 4000A from 1GHz to 1.5GHz really need a hardware mod?
datasheet says upgrade ay service center but I wonder what could be the difference between 4104A and 4154A front ends?

[EEVblog]

Does upgrading 4000A from 1GHz to 1.5GHz really need a hardware mod?
datasheet says upgrade ay service center but I wonder what could be the difference between 4104A and 4154A front ends?

My bet would be it's just better characterisation, probably with some extra software compensation points.
The upgrade is probably them just hooking it to their custom test system, it sweeps it and then stores the compensation results internally.

[analogRF]

Does upgrading 4000A from 1GHz to 1.5GHz really need a hardware mod?
datasheet says upgrade ay service center but I wonder what could be the difference between 4104A and 4154A front ends?

My bet would be it's just better characterisation, probably with some extra software compensation points.
The upgrade is probably them just hooking it to their custom test system, it sweeps it and then stores the compensation results internally.

so only a fresh calibration (alignment)?
it's a bit hard to believe they didnt calibrate it all the way to 1.5GHz if the board is designed for both 1GHz and 1.5GHz models like their lower freq models (350M to 500M for example etc...)

[TheSteve]

I have forgotten about this.
Does the 3000T PCB have a dedicated path layout for the 1GHz bandwidth input and the (132?!) components just left out?
And the 500MHz version uses another path that gets left unpopulated if it's the 1GHz model?

All of the 3000T bandwidth options use the same PCB. If you get the 500 MHz or lower version they use a different(lower cost) relay/attenuator combination. So as you guessed same PCB, but different paths/parts based on bandwidth.

[TheSteve]

Does upgrading 4000A from 1GHz to 1.5GHz really need a hardware mod?
datasheet says upgrade ay service center but I wonder what could be the difference between 4104A and 4154A front ends?

Yes, it will be subtle but different hardware. The lowpass filter between the front end amp and the ADC will be tuned differently.

[BillCRM]

Some info I acquired these years：
2000/3000/4000A/T/G shared the same input chip (code name llama, used in some late dso5000/6000 too), the same ADC (code name talon), and the ADC might be a slightly modified vesion from old dso5000/6000 series, and the waveform procseeor chip (Megazoom, code name baldwin, it also used in 6000X series too).
The 3000A can mod to 500Mhz, but can't go to 1GHz due to lack some pads.
The 3000T/G can mod to 1GHz, but need 4 expensive relays. I belive use some normal relays and external attenuation networks will also works.
The 4034A can software hack to 4054A. The hardware mod to 1/1.5GHz is basicly impossible. As the mod requires to install a fpga between the ADCs to do USB decode, and an agilent 1NE5-0001B is also need to finish the signal path. But if ignore the potential error message, only do the mod to input part, the mod will be very similar to 3000 series.
As for the calibration after those hardware mod, the factory cal is also possible by just add a command in the infiniivison.lnk. Then the user cal will magically turned into factory cal.
As for the serial number and model number hack, the agilent secure data tools (DTD,OPD) is also leaked. Be careful as there were already some modded machines on the market.
As for the Keygen, yes the Keygen, is also achieved by someone in China, but he won't release it of course. But hacked firmware is enough for us to use so no keygen is also not a big problem.
Hope these info will help someone in the future.

[TinLethax]

Some info I acquired these years：
2000/3000/4000A/T/G shared the same input chip (code name llama, used in some late dso5000/6000 too), the same ADC (code name talon), and the ADC might be a slightly modified vesion from old dso5000/6000 series, and the waveform procseeor chip (Megazoom, code name baldwin, it also used in 6000X series too).
The 3000A can mod to 500Mhz, but can't go to 1GHz due to lack some pads.
The 3000T/G can mod to 1GHz, but need 4 expensive relays. I belive use some normal relays and external attenuation networks will also works.
The 4034A can software hack to 4054A. The hardware mod to 1/1.5GHz is basicly impossible. As the mod requires to install a fpga between the ADCs to do USB decode, and an agilent 1NE5-0001B is also need to finish the signal path. But if ignore the potential error message, only do the mod to input part, the mod will be very similar to 3000 series.
As for the calibration after those hardware mod, the factory cal is also possible by just add a command in the infiniivison.lnk. Then the user cal will magically turned into factory cal.
As for the serial number and model number hack, the agilent secure data tools (DTD,OPD) is also leaked. Be careful as there were already some modded machines on the market.
As for the Keygen, yes the Keygen, is also achieved by someone in China, but he won't release it of course. But hacked firmware is enough for us to use so no keygen is also not a big problem.
Hope these info will help someone in the future.

No wonder why 2nd hand Agilent/Keysight scopes are so pricey on Taobao  .