added measure frequency and Mk2 libraries by TomQD-94 · Pull Request #54 · epics-modules/xspress3

TomQD-94 · 2024-05-28T10:27:51Z

This is a branch to allow using the community EPICS with Mk2 hardware.

Measures frequency of the clock to generate the right time frame for the acquisition time.

Updates to libraries for updates to firmware.

newville · 2024-07-26T15:36:50Z

@TomQD-94 Thanks - the timing parts look okay to me.

I'm not sure what the MOD_IMAGE and MOD_IMAGE3D changes mean. What are these for?

newville · 2024-07-30T19:28:47Z

@TomQD-94 Is this ready for merge?

TomQD-94 · 2024-08-06T08:44:40Z

@TomQD-94 Is this ready for merge?

I think now it is ready but would prefer a test from someone else to ensure I haven't forgotten something. Perhaps when I have confirmed it works with the shipped systems at your facility we can merge it in.

remove superfluous logging

… Mk2

fix of timing

Feature/ttl internal. TTL start the frame with provided frame length from itfg

* bzip-devel to bzip2-devel * erase maximum of requested (next run) and actual (last run) frames, avoiding errorenous timing values * disable circular buffers for GSE ioc * tweak GSE-7elem config * trim scalars saved by default via XML file, but add 'full' XML config * add short (25ms) sleep before checking if frame 0 acquire is done * Bring the iocDualMini_4ch IOC up-to-date, using the common startup files --------- Co-authored-by: Matt Newville <newville@cars.uchicago.edu> Co-authored-by: Tejas Guruswamy <tguruswamy@anl.gov>

BenBradnick

Looks good, just needs a bit of a tidy up to not lose the default configuration for IOCs.

iocs/xspress3IOC/iocBoot/common/SetMainValues.cmd

+dbpf("$(PREFIX)det1:NumImages",    1000)
+dbpf("$(PREFIX)det1:AcquireTime",  1)


iocs/xspress3IOC/iocBoot/ioc_2Channel/st.cmd

 # dbpf("$(PREFIX)det1:CONFIG_PATH", "/etc/xspress3/calibration/initial")
-dbpf("$(PREFIX)det1:CONFIG_PATH", "$(SUPPORT)/../xspress3_settings/current/")
+dbpf("$(PREFIX)det1:CONFIG_PATH", "/home/xspress3/xspress3_settings/4channel")



iocs/xspress3IOC/iocBoot/ioc_4Channel/st.cmd


 # Number of xspress3 cards and IP ADDR
-epicsEnvSet("XSP3CARDS", "1")
+epicsEnvSet("XSP3CARDS", "2")


iocs/xspress3IOC/iocBoot/ioc_4Channel/st.cmd

 #     card0_clocks.dat, chan1_reion0.dat, etc
 # dbpf("$(PREFIX)det1:CONFIG_PATH", "/etc/xspress3/calibration/initial")
-dbpf("$(PREFIX)det1:CONFIG_PATH", "$(SUPPORT)/../xspress3_settings/current/")
+dbpf("$(PREFIX)det1:CONFIG_PATH", "$(SUPPORT)/../xspress3_settings/4channel/")


xspress3App/opi/adl/xspress3_mcarois.adl

 	}
 	plotcom {
-		title="Last Frame"
+		title="Last Fram"


xspress3App/src/xspress3Epics.cpp

-      }
-
-      printf("xsp3_clocks_setup: Measured frequency %.2f MHz\n", float(xsp3_status)/1.0e6);
+    // TODO: XSP4_CLK_SRC_MIDPLN_LMK61E2 for Mk2 systems 


xspress3App/src/xspress3Epics.cpp

+    xsp3_status = xsp3->clocks_setup(xsp3_handle_, -1, generation == 3 ? XSP3M_CLK_SRC_LMK61E2 : (generation == 2 ? ((mark == 2) ? XSP4_CLK_SRC_MIDPLN_LMK61E2 : XSP3M_CLK_SRC_CDCM61004) : XSP3_CLK_SRC_XTAL),
+                                    XSP3_CLK_FLAGS_MASTER | XSP3_CLK_FLAGS_NO_DITHER, 0);
+    if (xsp3_status < 0) {


xspress3App/src/xspress3Epics.cpp

+    // TODO Is this needed for Mk2?
+    // u_int32_t actual_trigger_mode;
+    // xsp3_get_glob_timeA(xsp3_handle_, card, &actual_trigger_mode);
+    // }
+    // int xsp3_status = xsp3->set_sync_mode(xsp3_handle_, XSP3_SYNC_MIDPLANE, 0 , 0);
+    // if (xsp3_status != XSP3_OK) {
+    //   checkStatus(xsp3_status, "xsp3_set_sync_mode", functionName);
+    //   status = asynError;
    }


BenBradnick · 2025-09-10T09:23:48Z

@TomQD-94 Thanks - the timing parts look okay to me.

I'm not sure what the MOD_IMAGE and MOD_IMAGE3D changes mean. What are these for?

Almost all of the changes are just renaming data_float to data_type and changing some types from int to size_t. There are a couple of new functions added but I don't think these are used in the current Xspress library from a quick check.

The data module library is a generic library so it is used in many device libraries.

TomQD-94 · 2025-09-10T09:32:11Z

Branch ready for merge.
Adds TTL start internal frame trigger mode,
Adds Mk2 and Mk1 compatability with clock frequency check
also removes repeated called noted in issue. Resolves #56

TomQD-94 · 2025-09-10T13:54:42Z

@newville Please review and suggest any changes you want to this before merging in. Thanks

newville · 2025-09-11T03:11:58Z

@TomQD-94 Will this solve #56?

I added some workarounds in the code (and would have merged the related #62).

Do you think those added sleeps are no longer needed?

It would be really good to know how these things could be better tested with various Xspress3 generations.

BenBradnick · 2025-09-11T09:08:59Z

Hi @newville I have commented on your issue (#60) - I was unable to reproduce it here. If you still get the issue when testing this branch (and commenting out your original sleep) then I don't think #62 will cause any issues for systems which don't have the problem.

#56 should be resolved as the repeated call has been removed and we haven't identified any problems during testing so far.

newville · 2025-09-17T21:50:18Z

@TomQD-94 @BenBradnick OK, as discussed at #60, I think we should merge this, and I'll see if something specific to "very old Xspress3" requires the "sleep" discussed there and at #56.

TomQD-94 added 3 commits May 28, 2024 11:09

added measure frequency and Mk2 libraries

1191fd7

update to headers

f15f7b7

change header file unlink function

48b4ade

Check Mark2-ness and apply approprite clock source and timing

80dce97

TomQD-94 and others added 15 commits October 3, 2024 15:07

Mk1 clock fix

a731ba8

fix redefining variable

dacd452

Update xspress3Epics.cpp

999cd54

remove superfluous logging

cosmetic changes

8dfb078

rebase erge branch 'Mk2' of https://github.com/TomQD-94/xspress3 into…

2b73c23

… Mk2

fix of timing

0e1c484

tidying

aad9728

Merge pull request #2 from TomQD-94/Mk2_tidy

b4db66c

fix of timing

Adding extra trigger option

f1df966

adding trigger options to xspress header

304f33f

TTL + internal added

a701081

defining in the cpp file instead of the header

ba2bb97

Merge pull request #3 from TomQD-94/feature/TTL_internal

71a6dd6

Feature/ttl internal. TTL start the frame with provided frame length from itfg

Merge branch 'epics-modules:master' into Mk2

e8a0e43

TomQD-94 marked this pull request as ready for review September 9, 2025 09:59

BenBradnick suggested changes Sep 9, 2025

View reviewed changes

Tidied up per comments

befe79f

TomQD-94 requested a review from BenBradnick September 9, 2025 15:04

BenBradnick approved these changes Sep 10, 2025

View reviewed changes

directly setting XSP3READOUT from run_xspress3 when generated

fca4561

only update XSP3READOUT if it is unset

bb655ec

BenBradnick mentioned this pull request Sep 12, 2025

getNumFrames sets Frame number from 0 to 1 before first frame finishes #60

Open

newville merged commit 5928550 into epics-modules:master Sep 17, 2025

newville mentioned this pull request Sep 19, 2025

a better solution: wait at frame0 until getNumFramesRead() returns 0 #62

Open

		dbpf("$(PREFIX)det1:NumImages", 1000)
		dbpf("$(PREFIX)det1:AcquireTime", 1)

Conversation

TomQD-94 commented May 28, 2024

Uh oh!

newville commented Jul 26, 2024

Uh oh!

newville commented Jul 30, 2024

Uh oh!

TomQD-94 commented Aug 6, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

BenBradnick left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

BenBradnick commented Sep 10, 2025

Uh oh!

TomQD-94 commented Sep 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

TomQD-94 commented Sep 10, 2025

Uh oh!

newville commented Sep 11, 2025

Uh oh!

BenBradnick commented Sep 11, 2025

Uh oh!

newville commented Sep 17, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

TomQD-94 commented Aug 6, 2024 •

edited

Loading

TomQD-94 commented Sep 10, 2025 •

edited

Loading