Low Voltage LB status

Low Voltage LB status

RPC detector and trigger group

Pigi Paolucci - CMS LV review 2

Equipment type Total installed spare % of spareSY1527 SY1527 5 4 1 25,0%

A3485 A3485 5 3 2 66,7%

A3512N HV 127 116 11 9,5%

A1676 BC A1676 20 17 3 17,6%

EASY3000S CRATE 40 38 2 5,3%

EASY3000 CRATE 44 39 5 12,8%

A3000FB FAN 27 23 4 17,4%

A3486S_380 MAO AC/DC 12 10 2 20,0%

A3009 LV DC/DC 105 96 9 9,4%

A3801A Temperature AC/DC 17 14 3 21,4%RPC HV distributor Barrel HV Distribution Box 37 35 2 5,7%A3486S_230 Endcap MAO AC/DC 13 12 1 8,3%RPC HV distributor Endcap HV Distribution Box 24 24 0 0,0%

A3016 LV DC/DC 68 64 4 6,3%

RPC LinkBoardBox LBB 105 96 9 9,4%

RPC MasterBoard MLB 351 300 49 16,3%

RPC SlaveBoard SLB 574 500 70 14,0%

RPC ControlBoard CB 124 120 4 3,3%

RPC LB FrontPanel LB FP 125 120 5 4,2%

RPC OptoBox OB 14 10 4 40,0%

7/6/10

Pigi Paolucci - CMS LV review 3

Low Voltage Link Board status

• After many failures in the link board system the RPC-Trigger community decided to investigate all the possible causes analyzing not only the LB system but also the LV system:

• A3016 boards• MAO• LV and MAO power cut and power cycle• Power off and on of the link box• CAEN firmware

7/6/10

A3016 behavior cutting the MAO 3Phase

Pigi Paolucci - CMS LV review 5

LV link board Schema

1. MAO 48 service from BC– Output current = 15-18 A

2. EASY 3000 – Joint 48 Power & Service

3. Output power from A3016:– 2.5 Volts / 5.0 Amp– 3.5 Volts / 10.0 Amp

MAO

SY1527

48 service

3-phase

EASYA3016

48 service48 power

Link Box7/6/10

Pigi Paolucci - CMS LV review 6

What powering OFF the MAO in the proper ways ?

• Powering OFF the MAO via SY1527 or using the interlock placed on the front

OK

7/6/10

Pigi Paolucci - CMS LV review 7

What powering OFF the EASY in a “bad” way ?

• Powering OFF the EASY 3000 crate removing the power or service connector from the back

OK

7/6/10

Pigi Paolucci - CMS LV review 8

What cutting the MAO 3-phase ?

• Shutting-down the MAO cutting 3-phase

•oscillation does not depend from the MAO load (from 4 A to 18 A)

•sometimes 2 oscillations in few msec

•oscillation disappear separating the 48 power from the 48 service

•No other ways to reproduce it

NOT OK7/6/10

Pigi Paolucci - CMS LV review 9

What happen cutting MAO 3-phase ?• MAO goes OFF in 2 steps: from 48 to 22 V and a from 22 to 0 V • The A3016 channels go off when MAO voltage is around 22 V and

this takes about 250-300 msec (fixed load on the MAO)• Around 22 V the boards begin to work not properly (Vcc goes from

12 V to 6 V) and generates oscillations on Vout.

7/6/10

Vmao = 22 VVcc = 12 V

6 V

7/6/10 9Pigi Paolucci - INFN

Mao behavior during 3Phase glitch

Pigi Paolucci - CMS LV review 11

T0 T1

At T0:• Turn off the 3-phase switch• MAO starts ramping down with a “dummy load” - NO status changing • After a while, it gives UNV status, but the channel keeps status ON.

At T1:• MAO ramps down with different slope (real load)

At T2:• MAO swiched ON (600 ms from T0).

Switching OFF-ON in <1sec, no IntFail status is generated by the MAO and the channel goes back to the 48V without any alarm acknowledge.

T2

Scope output

MAO

A3016

40 V

7/6/10

Pigi Paolucci - CMS LV review 12

What we have done !Already implemented Solutions:

• Introducing a big capacitor (33 mF) between the service and power line of the EASY

• Improve the A3016 firmware • Modify the MAO firmware

Long Term Solutions:1. Designing a protection on the Link Box LV lines

(see tomorrow meeting) 2. Separating the 48 service and power of the

EASY ?7/6/10

OK

Under test now

Under test now

Pigi Paolucci - CMS LV review 13

A3016 firmware requirements• Couple of channels in tracking mode (ON/OFF/Trip at same time

and with one command).– delay between the 2 channels must be less than 1 msec

• Ramping up in stepping mode:– overshoot first steps < 10% of the setting voltage– overshoot of the last step must be < setting voltage– voltage must be always < setting voltage (no overshoot above max value)– Over-voltages are very dangerous for the LB.

– no under-voltage dips are allowed after VCC2.5 gets above 2.0 V– no under-voltage dips are allowed after VCC3.3 gets above 1.2 V– total time to reach the setting voltages (3.5 V and 2.5 V) should be less than 60

msec

7/6/10

Pigi Paolucci - CMS LV review 14

MAO firmware requirements

• The MAO behavior is case of glitch on the 3-phase line is potentially dangerous for the LV system.

• Turn OFF the MAO when the Vmon < 43-40 Volts in about 100 msec and require an acknowledge afterwards to bring the channel back to ON.– In 100 msec firmware measures the voltage 5 times,

should be enough.– After 100 msec the MAO voltage is around 35-38 V

and it is enough to shut-down properly the A3016.

7/6/10

Pigi Paolucci - CMS LV review 157/6/10

MAO

Vcc

Vout

New MAO firmware

old MAO firmware

Oscillations on both Vccand Vout

200 msecOFF – 38V

43V

New MOA firmware

Pigi Paolucci - CMS LV review 167/6/10

Rumping up - new A3016 firmware

50 msec

80 msec

2.5 V

3.5 V

No overshot over the max voltage

Pigi Paolucci - CMS LV review 177/6/10

Overshot - new A3016 firmware

Overshot < 10-15% voltage set

Pigi Paolucci - CMS LV review 187/6/10

Channels delay- new A3016 firmware

2 msec

Delay stable and < 2 msec (depends on the charge)

Pigi Paolucci - CMS LV review 19

Results with C and diode

power

service

1.5 sec

7/6/10

Pigi Paolucci - CMS LV review 20

Results with C and diode

2.5 V

48 VNo spikes or oscillations

600 msec

7/6/10

Pigi Paolucci - CMS LV review 21

Conclusions• Capacitor protection is working safe

• New firmwares are ready and under test at CAEN.• Next week will be tested at 904• Both Capacitor boxes and firmwares have been

payed by INFN (M&O-B 2010) 16 Keuro

• Hardware protection of the Link Box is in any case necessary

7/6/10