UmTRXv2 final project

Jean-Samuel Najnudel - BJT PARTNERS SARL jsn at
Tue Jul 3 20:36:42 UTC 2012

Hi Alexander,

On 7/3/12, Alexander Chemeris <alexander.chemeris at> wrote:
> Hi Jean-Samuel,
> On Tue, Jul 3, 2012 at 10:59 PM, Jean-Samuel Najnudel - BJT PARTNERS
> SARL <jsn at> wrote:
>> Hi Alexander,
>> Thank you very much for your e-mail.
>> 1/ They did not started the PCB manufacturing yet. They are still
>> waiting for my feedback on the questions I asked to Andrey.
>> Actually, if we need to replace the SRAM and/or the flash to use
>> cheaper equivalent components, we may need to reroute the PCB. This is
>> why I ask them to wait for the feedback from Andrey before starting
>> the PCB production. They can start as soon as we let them know what we
>> do about the SRAM and the flash (with or without PCB files update).
>> Anyway, I was not very worried about the delay. As discussed together
>> last time in Paris, before deploying the UmTRX on the field, we would
>> need OpenBTS to support handover. Even for tests, if I open the
>> network without handover, because of the terrain in Mayotte, there
>> will be so much call drops nobody in Mayotte will accept to use our
>> network anymore in the future. Really, handover is very important in
>> Mayotte.
>> As I understand OpenBTS is under development but it is quite hard
>> work. It will probably not be ready before the end of this summer.
>> This is why I did not feel very worried about the PCB production
>> delay.
> We really need to get prototypes ASAP for many reasons - we have to
> start certification process, we still need to find enclosures, we have
> other pilots pending, etc. And every week of delay is painful for us.

Ok. I will ask the fab to start the production ASAP.

>> Anyway, I will do my best to get the UmTRX prototypes ready as soon as
>> possible. I just need a feedback from Andrey because, if it does not
>> lower the UmTRX board performances, we should really try to find an
>> alternative solution to use cheaper SRAM and flash chips. This could
>> improve a lot the cost structure of the UmTRX.
> We should be fine with this chip:
> (with no PCB changes).

This looks good. However, this SRAM is 18 MBits instead of 36 MBits.
This makes a quite big difference.
Are you sure this will be fine ?
Andrey, do you confirm this is fine ? Is this chip fully pin compatible ?

>> 2/ I did not contacted Lime yet but I will do this as soon as
>> possible. I will do my best to get the LMS chips on time.
>> By the way, for both PCB assembly and LMS order, how many UmTRX board
>> would you like the fab makes ? We discussed together about 5 or 10
>> units but we did not took the final decision yet. It can be any
>> reasonable number between 5 and 10. Please let me know the quantity
>> you would need. I will ask the fab and order the component quantities
>> we need.
> We have only 5 FPGAs left, so I think we should assembly 5 prototypes.
> I.e. we need 10 LMS chips.

This is fine for me. I will do this.

> But we could safely order more PCBs, because we're pretty sure PCB won't
> change.

Yes, for sure I will. You are completly right. I asked for the price.
There is only about 100 Euros difference in total between 5 and 20
Moreover, ordering 20 PCB will save us a lot on next batches. It will
also save us a lot of valuable time if we need some more prototypes.
Actually, the PCB is what makes the production quite long.

Anyway, I ask the fab will move toward the production asap.

Best regards.


>> By the way, the meeting with the fab this morning was fine and they
>> are ready to start the production asap.
> Please, start PCB manufacturing ASAP.
> --
> Regards,
> Alexander Chemeris.
> CEO, Fairwaves LLC / ООО УмРадио

More information about the UmTRX mailing list